Rachel C. Hill, MD

Use of artificial intelligence (AI) in dermatology has increased over the past decade, likely driven by advances in deep learning algorithms, computing hardware, and machine learning.¹ Studies comparing the performance of AI algorithms to dermatologists in classifying skin disorders have shown conflicting results.^2,3 In this study, we aimed to analyze AI tools used for diagnosing and classifying skin disease and evaluate the role of dermatologists in the creation of AI technology. We also investigated the number of clinical images used in datasets to train AI programs and compared tools that were created with dermatologist input to those created without dermatologist/clinician involvement.

Methods

A search of PubMed articles indexed for MEDLINE using the terms machine learning, artificial intelligence, and dermatology was conducted on September 18, 2022. Articles were included if they described full-length trials; used machine learning for diagnosis of or screening for dermatologic conditions; and used dermoscopic or gross image datasets of the skin, hair, or nails. Articles were categorized into 4 groups based on the conditions covered: chronic wounds, inflammatory skin diseases, mixed conditions, and pigmented skin lesions. Algorithms were sorted into 4 categories: convolutional/convoluted neural network, deep learning model/deep neural network, AI/artificial neural network, and other. Details regarding Fitzpatrick skin type and skin of color (SoC) inclusion in the articles or AI algorithm datasets were recorded. Univariate and multivariate analyses were performed using Microsoft Excel and SAS Studio 3.8. Sensitivity and specificity were calculated for all included AI technology. Sensitivity, specificity, and the number of clinical images were compared among the included articles using analysis of variance and t tests (α=0.05; P<.05 indicated statistical significance).

Results

Our search yielded 1016 articles, 58 of which met the inclusion criteria. Overall, 25.9% (15/58) of the articles utilized AI to diagnose or classify mixed skin diseases; 22.4% (13/58) for pigmented skin lesions; 19.0% (11/58) for wounds; 17.2% (10/58) for inflammatory skin diseases; and 5.2% (3/58) each for acne, psoriasis, and onychomycosis. Overall, 24.0% (14/58) of articles provided information about Fitzpatrick skin type, and 58.7% (34/58) included clinical images depicting SoC. Furthermore, we found that only 20.7% (12/58) of articles on deep learning models included descriptions of patient ethnicity or race in at least 1 dataset, and only 10.3% (6/58) of studies included any information about skin tone in the dataset. Studies with a dermatologist as the last author (most likely to be supervising the project) were more likely to include clinical images depicting SoC than those without (82.6% [19/23] and 16.7% [3/18], respectively [P=.0411]).

The mean (SD) number of clinical images in the study articles was 28,422 (84,050). Thirty-seven (63.8%) of the study articles included gross images, 17 (29.3%) used dermoscopic images, and 4 (6.9%) used both. Twenty-seven (46.6%) articles used convolutional/convoluted neural networks, 15 (25.9%) used deep learning model/deep neural networks, 8 (13.8%) used other algorithms, 6 (10.3%) used AI/artificial neural network, and 2 (3.4%) used fuzzy algorithms. Most studies were conducted in China (29.3% [17/58]), Germany (12.1% [7/58]), India (10.3% [6/58]), multiple nations (10.3% [6/58]), and the United States (10.3% [6/58]). Overall, 82.8% (48/58) of articles included at least 1 dermatologist coauthor. Sensitivity of the AI models was 0.85, and specificity was 0.85. The average percentage of images in the dataset correctly identified by a physician was 76.87% vs 81.62% of images correctly identified by AI. Average agreement between AI and physician assessment was 77.98%, defined as AI and physician both having the same diagnosis. 

Articles authored by dermatologists contained more clinical images than those without dermatologists in key authorship roles (P<.0001)(eTable). Psoriasis-related algorithms had the fewest (mean [SD]: 3173 [4203]), and pigmented skin lesions had the most clinical images (mean [SD]: 53,19l [155,579]).

Comment

Our results indicated that AI studies with dermatologist authors had significantly more images in their datasets (ie, the set of clinical images of skin lesions used to train AI algorithms in diagnosing or classifying lesions) than those with nondermatologist authors (P<.0001)(eTable). Similarly, in a study of AI technology for skin cancer diagnosis, AI studies with dermatologist authors (ie, included in the development of the AI algorithm) had more images than studies without dermatologist authors.¹ Deep learning textbooks have suggested that 5000 clinical images or training input per output category are needed to produce acceptable algorithm performance, and more than 10 million are needed to produce results superior to human performance.^4-10 Despite advances in AI for dermatologic image analysis, the creation of these models often has been directed by nondermatologists¹; therefore, dermatologist involvement in AI development is necessary to facilitate collection of larger image datasets and optimal performance for image diagnosis/classification tasks.

We found that 20.7% of articles on deep learning models included descriptions of patient ethnicity or race, and only 10.3% of studies included any information about skin tone in the dataset. Furthermore, American investigators primarily trained models using clinical images of patients with lighter skin tones, whereas Chinese investigators exclusively included images depicting darker skin tones. Similarly, in a study of 52 cutaneous imaging deep learning articles, only 17.3% (9/52) reported race and/or Fitzpatrick skin type, and only 7.7% (4/52) of articles included both.^2,6,8 Therefore, dermatologists are needed to contribute images representing diverse populations and collaborate in AI research studies, as their involvement is necessary to ensure the accuracy of AI models in classifying lesions or diagnosing skin lesions across all skin types.

Our search was limited to PubMed, and real-world applications could not be evaluated.

Conclusion

In summary, we found that AI studies with dermatologist authors used larger numbers of clinical images in their datasets and more images representing diverse skin types than studies without. Therefore, we advocate for greater involvement of dermatologists in AI research, which might result in better patient outcomes by improving diagnostic accuracy.

Use of artificial intelligence (AI) in dermatology has increased over the past decade, likely driven by advances in deep learning algorithms, computing hardware, and machine learning.¹ Studies comparing the performance of AI algorithms to dermatologists in classifying skin disorders have shown conflicting results.^2,3 In this study, we aimed to analyze AI tools used for diagnosing and classifying skin disease and evaluate the role of dermatologists in the creation of AI technology. We also investigated the number of clinical images used in datasets to train AI programs and compared tools that were created with dermatologist input to those created without dermatologist/clinician involvement.

Methods

A search of PubMed articles indexed for MEDLINE using the terms machine learning, artificial intelligence, and dermatology was conducted on September 18, 2022. Articles were included if they described full-length trials; used machine learning for diagnosis of or screening for dermatologic conditions; and used dermoscopic or gross image datasets of the skin, hair, or nails. Articles were categorized into 4 groups based on the conditions covered: chronic wounds, inflammatory skin diseases, mixed conditions, and pigmented skin lesions. Algorithms were sorted into 4 categories: convolutional/convoluted neural network, deep learning model/deep neural network, AI/artificial neural network, and other. Details regarding Fitzpatrick skin type and skin of color (SoC) inclusion in the articles or AI algorithm datasets were recorded. Univariate and multivariate analyses were performed using Microsoft Excel and SAS Studio 3.8. Sensitivity and specificity were calculated for all included AI technology. Sensitivity, specificity, and the number of clinical images were compared among the included articles using analysis of variance and t tests (α=0.05; P<.05 indicated statistical significance).

Results

Our search yielded 1016 articles, 58 of which met the inclusion criteria. Overall, 25.9% (15/58) of the articles utilized AI to diagnose or classify mixed skin diseases; 22.4% (13/58) for pigmented skin lesions; 19.0% (11/58) for wounds; 17.2% (10/58) for inflammatory skin diseases; and 5.2% (3/58) each for acne, psoriasis, and onychomycosis. Overall, 24.0% (14/58) of articles provided information about Fitzpatrick skin type, and 58.7% (34/58) included clinical images depicting SoC. Furthermore, we found that only 20.7% (12/58) of articles on deep learning models included descriptions of patient ethnicity or race in at least 1 dataset, and only 10.3% (6/58) of studies included any information about skin tone in the dataset. Studies with a dermatologist as the last author (most likely to be supervising the project) were more likely to include clinical images depicting SoC than those without (82.6% [19/23] and 16.7% [3/18], respectively [P=.0411]).

The mean (SD) number of clinical images in the study articles was 28,422 (84,050). Thirty-seven (63.8%) of the study articles included gross images, 17 (29.3%) used dermoscopic images, and 4 (6.9%) used both. Twenty-seven (46.6%) articles used convolutional/convoluted neural networks, 15 (25.9%) used deep learning model/deep neural networks, 8 (13.8%) used other algorithms, 6 (10.3%) used AI/artificial neural network, and 2 (3.4%) used fuzzy algorithms. Most studies were conducted in China (29.3% [17/58]), Germany (12.1% [7/58]), India (10.3% [6/58]), multiple nations (10.3% [6/58]), and the United States (10.3% [6/58]). Overall, 82.8% (48/58) of articles included at least 1 dermatologist coauthor. Sensitivity of the AI models was 0.85, and specificity was 0.85. The average percentage of images in the dataset correctly identified by a physician was 76.87% vs 81.62% of images correctly identified by AI. Average agreement between AI and physician assessment was 77.98%, defined as AI and physician both having the same diagnosis. 

Articles authored by dermatologists contained more clinical images than those without dermatologists in key authorship roles (P<.0001)(eTable). Psoriasis-related algorithms had the fewest (mean [SD]: 3173 [4203]), and pigmented skin lesions had the most clinical images (mean [SD]: 53,19l [155,579]).

Comment

Our results indicated that AI studies with dermatologist authors had significantly more images in their datasets (ie, the set of clinical images of skin lesions used to train AI algorithms in diagnosing or classifying lesions) than those with nondermatologist authors (P<.0001)(eTable). Similarly, in a study of AI technology for skin cancer diagnosis, AI studies with dermatologist authors (ie, included in the development of the AI algorithm) had more images than studies without dermatologist authors.¹ Deep learning textbooks have suggested that 5000 clinical images or training input per output category are needed to produce acceptable algorithm performance, and more than 10 million are needed to produce results superior to human performance.^4-10 Despite advances in AI for dermatologic image analysis, the creation of these models often has been directed by nondermatologists¹; therefore, dermatologist involvement in AI development is necessary to facilitate collection of larger image datasets and optimal performance for image diagnosis/classification tasks.

We found that 20.7% of articles on deep learning models included descriptions of patient ethnicity or race, and only 10.3% of studies included any information about skin tone in the dataset. Furthermore, American investigators primarily trained models using clinical images of patients with lighter skin tones, whereas Chinese investigators exclusively included images depicting darker skin tones. Similarly, in a study of 52 cutaneous imaging deep learning articles, only 17.3% (9/52) reported race and/or Fitzpatrick skin type, and only 7.7% (4/52) of articles included both.^2,6,8 Therefore, dermatologists are needed to contribute images representing diverse populations and collaborate in AI research studies, as their involvement is necessary to ensure the accuracy of AI models in classifying lesions or diagnosing skin lesions across all skin types.

Our search was limited to PubMed, and real-world applications could not be evaluated.

Conclusion

In summary, we found that AI studies with dermatologist authors used larger numbers of clinical images in their datasets and more images representing diverse skin types than studies without. Therefore, we advocate for greater involvement of dermatologists in AI research, which might result in better patient outcomes by improving diagnostic accuracy.

Use of artificial intelligence (AI) in dermatology has increased over the past decade, likely driven by advances in deep learning algorithms, computing hardware, and machine learning.¹ Studies comparing the performance of AI algorithms to dermatologists in classifying skin disorders have shown conflicting results.^2,3 In this study, we aimed to analyze AI tools used for diagnosing and classifying skin disease and evaluate the role of dermatologists in the creation of AI technology. We also investigated the number of clinical images used in datasets to train AI programs and compared tools that were created with dermatologist input to those created without dermatologist/clinician involvement.

Methods

A search of PubMed articles indexed for MEDLINE using the terms machine learning, artificial intelligence, and dermatology was conducted on September 18, 2022. Articles were included if they described full-length trials; used machine learning for diagnosis of or screening for dermatologic conditions; and used dermoscopic or gross image datasets of the skin, hair, or nails. Articles were categorized into 4 groups based on the conditions covered: chronic wounds, inflammatory skin diseases, mixed conditions, and pigmented skin lesions. Algorithms were sorted into 4 categories: convolutional/convoluted neural network, deep learning model/deep neural network, AI/artificial neural network, and other. Details regarding Fitzpatrick skin type and skin of color (SoC) inclusion in the articles or AI algorithm datasets were recorded. Univariate and multivariate analyses were performed using Microsoft Excel and SAS Studio 3.8. Sensitivity and specificity were calculated for all included AI technology. Sensitivity, specificity, and the number of clinical images were compared among the included articles using analysis of variance and t tests (α=0.05; P<.05 indicated statistical significance).

Results

Our search yielded 1016 articles, 58 of which met the inclusion criteria. Overall, 25.9% (15/58) of the articles utilized AI to diagnose or classify mixed skin diseases; 22.4% (13/58) for pigmented skin lesions; 19.0% (11/58) for wounds; 17.2% (10/58) for inflammatory skin diseases; and 5.2% (3/58) each for acne, psoriasis, and onychomycosis. Overall, 24.0% (14/58) of articles provided information about Fitzpatrick skin type, and 58.7% (34/58) included clinical images depicting SoC. Furthermore, we found that only 20.7% (12/58) of articles on deep learning models included descriptions of patient ethnicity or race in at least 1 dataset, and only 10.3% (6/58) of studies included any information about skin tone in the dataset. Studies with a dermatologist as the last author (most likely to be supervising the project) were more likely to include clinical images depicting SoC than those without (82.6% [19/23] and 16.7% [3/18], respectively [P=.0411]).

The mean (SD) number of clinical images in the study articles was 28,422 (84,050). Thirty-seven (63.8%) of the study articles included gross images, 17 (29.3%) used dermoscopic images, and 4 (6.9%) used both. Twenty-seven (46.6%) articles used convolutional/convoluted neural networks, 15 (25.9%) used deep learning model/deep neural networks, 8 (13.8%) used other algorithms, 6 (10.3%) used AI/artificial neural network, and 2 (3.4%) used fuzzy algorithms. Most studies were conducted in China (29.3% [17/58]), Germany (12.1% [7/58]), India (10.3% [6/58]), multiple nations (10.3% [6/58]), and the United States (10.3% [6/58]). Overall, 82.8% (48/58) of articles included at least 1 dermatologist coauthor. Sensitivity of the AI models was 0.85, and specificity was 0.85. The average percentage of images in the dataset correctly identified by a physician was 76.87% vs 81.62% of images correctly identified by AI. Average agreement between AI and physician assessment was 77.98%, defined as AI and physician both having the same diagnosis. 

Articles authored by dermatologists contained more clinical images than those without dermatologists in key authorship roles (P<.0001)(eTable). Psoriasis-related algorithms had the fewest (mean [SD]: 3173 [4203]), and pigmented skin lesions had the most clinical images (mean [SD]: 53,19l [155,579]).

Comment

Our results indicated that AI studies with dermatologist authors had significantly more images in their datasets (ie, the set of clinical images of skin lesions used to train AI algorithms in diagnosing or classifying lesions) than those with nondermatologist authors (P<.0001)(eTable). Similarly, in a study of AI technology for skin cancer diagnosis, AI studies with dermatologist authors (ie, included in the development of the AI algorithm) had more images than studies without dermatologist authors.¹ Deep learning textbooks have suggested that 5000 clinical images or training input per output category are needed to produce acceptable algorithm performance, and more than 10 million are needed to produce results superior to human performance.^4-10 Despite advances in AI for dermatologic image analysis, the creation of these models often has been directed by nondermatologists¹; therefore, dermatologist involvement in AI development is necessary to facilitate collection of larger image datasets and optimal performance for image diagnosis/classification tasks.

We found that 20.7% of articles on deep learning models included descriptions of patient ethnicity or race, and only 10.3% of studies included any information about skin tone in the dataset. Furthermore, American investigators primarily trained models using clinical images of patients with lighter skin tones, whereas Chinese investigators exclusively included images depicting darker skin tones. Similarly, in a study of 52 cutaneous imaging deep learning articles, only 17.3% (9/52) reported race and/or Fitzpatrick skin type, and only 7.7% (4/52) of articles included both.^2,6,8 Therefore, dermatologists are needed to contribute images representing diverse populations and collaborate in AI research studies, as their involvement is necessary to ensure the accuracy of AI models in classifying lesions or diagnosing skin lesions across all skin types.

Our search was limited to PubMed, and real-world applications could not be evaluated.

Conclusion

In summary, we found that AI studies with dermatologist authors used larger numbers of clinical images in their datasets and more images representing diverse skin types than studies without. Therefore, we advocate for greater involvement of dermatologists in AI research, which might result in better patient outcomes by improving diagnostic accuracy.

The Diagnosis: Glomangiomyoma

The nail unit excision specimen showed collections of cuboidal cells and spindled cells within the corium that were consistent with a diagnosis of a glomangiomyoma, a rare glomus tumor variant (Figure). Glomus tumors are benign neoplasms comprising glomus bodies, which are arteriovenous anastomoses involved in thermoregulation.¹ They develop in areas densely populated by glomus bodies, including the fingers, toes, and subungual areas. Glomus tumors most commonly develop in middle-aged women.² Clinically, they manifest with a characteristic triad of intense pain, point tenderness, and cold sensitivity and may appear as reddish-pink or blue macules under the nail plate and/or longitudinal erythronychia.^2-6 The presence of multiple glomus tumors is associated with neurofibromatosis type 1.⁷ 

Advanced imaging including ultrasonography and magnetic resonance imaging (MRI) may help confirm the diagnosis but may not be cost effective, as excision with histopathology is needed to relieve symptoms and render a definitive diagnosis. Radiography is highly insensitive in identifying bone erosions associated with glomus tumors.⁸ With ultrasonography, glomus tumors appear hypoechoic; with Doppler ultrasonography, they appear hypervascular. With MRI, glomus tumors appear as well-defined nodular lesions with hypointense signal intensity on T1-weighted sequence and hyperintense signal intensity on T2-weighted sequence, with strong enhancement using gadolinium-based contrast.^9,10 On histopathology, a glomus tumor appears as a nodular tumor with sheets of oval-nucleated cells arranged in multicellular layers surrounding blood vessels and are immunoreactive for α-smooth muscle actin, muscle-specific actin, and type IV collagen.^11,12 

There are several glomus tumor variants. The most common is a solid glomus tumor, which predominantly is composed of glomus cells, followed by glomangioma, which mainly is composed of blood vessels. Glomangiomyoma, which mostly is composed of smooth muscle cells, is the rarest variant.¹³ 

While glomus tumors are common in the subungual areas, it is an uncommon location for glomangiomyomas, which have been reported in the nail unit in only 7 prior case reports identified through searches of PubMed and Google Scholar using the terms glomangiomyoma, glomangiomyoma nail, and subungual glomangiomyoma (Table).^13-19 Glomangiomyomas more commonly are described in solid organs, including the stomach, kidney, pancreas, and bladder.¹⁶ The mean age of patients with subungual glomangiomyomas, including our patient, was 40.4 years (range, 3-61 years), with the majority being female (75.0% [6/8]). Most patients presented with fingernail involvement (75.0% [6/8]), nail dystrophy (eg, nail plate thinning, longitudinal grooves, splinter hemorrhages, longitudinal erythronychia)(62.5% [5/8]), and intermittent pain and/or point tenderness in the affected nail (75.0% [6/8]).^13-19 Notably, only our patient had longitudinal erythronychia as a clinical feature, and only one other case described MRI findings, which included a lobulated mass with intense contrast and distal phalanx destruction.¹⁸ One patient was a 3-year-old girl with a family history of generalized multiple glomangiomyomas. Although subungual glomangiomyoma was not confirmed on histopathology, the diagnosis in this patient was presumed based on her family history.¹³ On histopathology, glomangiomyomas are composed of oval-nucleated cells surrounding blood vessels. These oval-nucleated cells then gradually transition to smooth muscle cells.²⁰ 

A myxoid cyst is composed of a pseudocyst, which lacks a cyst lining, and is a result of synovial fluid from the distal interphalangeal joint entering the pseudocyst space.² It typically manifests with a longitudinal groove in the nail plate. A flesh-colored nodule may be appreciated between the cuticle and the distal interphalangeal joint.² The depth of the longitudinal groove may vary depending on the volume of synovial fluid within the myxoid cyst.²¹ In a series of 35 cases of subungual myxoid cysts, none manifested with longitudinal erythronychia. Due to their composition, myxoid cysts can be distinguished easily from solid tumors of the nail unit via transillumination.²² Pain is a much less common with myxoid cysts vs glomus tumors, as the filling of the pseudocyst space with synovial fluid typically is gradual, allowing the surrounding tissue to accommodate and adapt over time.²¹ In equivocal cases, MRI or high-resolution ultrasonography may be used to distinguish myxoid cysts and glomus tumors.⁸ Histopathology shows accumulation of mucin in the dermis with surrounding fibrous stroma.²³

Subungual neuromas are painful benign tumors that develop due to disorganized neural proliferation following disruption to peripheral nerves secondary to trauma or surgery. In 3 case reports, subungual neuromas manifested as painful subungual nodules, with proximal nail plate ridging, or onycholysis.^24-26 Since neuromas have only rarely been described in the subungual region, reports of MRI and ultrasonography findings are unknown. Histopathology is needed to distinguish neuromas from glomus tumors. Histopathology shows an acapsular structure consisting of disorganized spindle-cell proliferation and nerve fibers arranged in a tangle of fascicles within fibrotic tissue.²⁵ On immunochemistry, spindle cells typically are positive for cellular antigen protein S100.²⁶ 

Leiomyomas are benign neoplasms derived from smooth muscle, typically localized to the uterus or gastrointestinal tract, and have been described rarely in the nail unit.^27,28 It is hypothesized that subungual leiomyomas originate from the vascular smooth muscle in the subcutaneous layer of the nail unit.²⁸ Like glomus tumors, leiomyomas of the subungual region often manifest with pain and longitudinal erythronychia.^27-30 Subungual leiomyomas may be distinguished from glomus tumors via advanced imaging techniques, including ultrasonography and MRI. Cutaneous leiomyomas have been described with mild to moderate internal low flow vascularity on Doppler ultrasonography, while glomus tumors typically reveal high internal vascularity.²⁸ Biopsy with histopathology is needed for definitive diagnosis. On histopathology, leiomyomas demonstrate bland-appearing spindle-shaped cells with elongated nuclei arranged in fascicles.²⁷ They typically are positive for α-smooth muscle actin and caldesmon on immunostaining. 

Eccrine spiradenomas are benign adnexal tumors likely of apocrine origin with limited case reports in the literature.^31,32 Clinically, eccrine spiradenomas involving the nail unit may manifest with longitudinal nail splitting of the nail or as a papule on the proximal nail fold, with associated tenderness.^31,32 In a report of a 50-year-old woman with a histopathologically confirmed eccrine spiradenoma manifesting with longitudinal splitting of the nail and pain in the proximal nail fold, the mass appeared hypoechoic on ultrasonography with increased intramass vascularity on Doppler, while MRI showed an intensely enhancing lesion.³¹ These imaging features, combined with a classically manifesting feature of pain, make eccrine spiradenomas difficult to distinguish from glomus tumors; therefore, histopathologic examination can provide a definitive diagnosis, and surgical excision is used for treatment.³¹ On histopathology, these tumors are well circumscribed and composed of both small dark basaloid cells with peripheral compact nuclei and larger cells with central pale nuclei, which may be arranged in tubules.^31,32

The Diagnosis: Glomangiomyoma

The nail unit excision specimen showed collections of cuboidal cells and spindled cells within the corium that were consistent with a diagnosis of a glomangiomyoma, a rare glomus tumor variant (Figure). Glomus tumors are benign neoplasms comprising glomus bodies, which are arteriovenous anastomoses involved in thermoregulation.¹ They develop in areas densely populated by glomus bodies, including the fingers, toes, and subungual areas. Glomus tumors most commonly develop in middle-aged women.² Clinically, they manifest with a characteristic triad of intense pain, point tenderness, and cold sensitivity and may appear as reddish-pink or blue macules under the nail plate and/or longitudinal erythronychia.^2-6 The presence of multiple glomus tumors is associated with neurofibromatosis type 1.⁷ 

Advanced imaging including ultrasonography and magnetic resonance imaging (MRI) may help confirm the diagnosis but may not be cost effective, as excision with histopathology is needed to relieve symptoms and render a definitive diagnosis. Radiography is highly insensitive in identifying bone erosions associated with glomus tumors.⁸ With ultrasonography, glomus tumors appear hypoechoic; with Doppler ultrasonography, they appear hypervascular. With MRI, glomus tumors appear as well-defined nodular lesions with hypointense signal intensity on T1-weighted sequence and hyperintense signal intensity on T2-weighted sequence, with strong enhancement using gadolinium-based contrast.^9,10 On histopathology, a glomus tumor appears as a nodular tumor with sheets of oval-nucleated cells arranged in multicellular layers surrounding blood vessels and are immunoreactive for α-smooth muscle actin, muscle-specific actin, and type IV collagen.^11,12 

There are several glomus tumor variants. The most common is a solid glomus tumor, which predominantly is composed of glomus cells, followed by glomangioma, which mainly is composed of blood vessels. Glomangiomyoma, which mostly is composed of smooth muscle cells, is the rarest variant.¹³ 

While glomus tumors are common in the subungual areas, it is an uncommon location for glomangiomyomas, which have been reported in the nail unit in only 7 prior case reports identified through searches of PubMed and Google Scholar using the terms glomangiomyoma, glomangiomyoma nail, and subungual glomangiomyoma (Table).^13-19 Glomangiomyomas more commonly are described in solid organs, including the stomach, kidney, pancreas, and bladder.¹⁶ The mean age of patients with subungual glomangiomyomas, including our patient, was 40.4 years (range, 3-61 years), with the majority being female (75.0% [6/8]). Most patients presented with fingernail involvement (75.0% [6/8]), nail dystrophy (eg, nail plate thinning, longitudinal grooves, splinter hemorrhages, longitudinal erythronychia)(62.5% [5/8]), and intermittent pain and/or point tenderness in the affected nail (75.0% [6/8]).^13-19 Notably, only our patient had longitudinal erythronychia as a clinical feature, and only one other case described MRI findings, which included a lobulated mass with intense contrast and distal phalanx destruction.¹⁸ One patient was a 3-year-old girl with a family history of generalized multiple glomangiomyomas. Although subungual glomangiomyoma was not confirmed on histopathology, the diagnosis in this patient was presumed based on her family history.¹³ On histopathology, glomangiomyomas are composed of oval-nucleated cells surrounding blood vessels. These oval-nucleated cells then gradually transition to smooth muscle cells.²⁰ 

A myxoid cyst is composed of a pseudocyst, which lacks a cyst lining, and is a result of synovial fluid from the distal interphalangeal joint entering the pseudocyst space.² It typically manifests with a longitudinal groove in the nail plate. A flesh-colored nodule may be appreciated between the cuticle and the distal interphalangeal joint.² The depth of the longitudinal groove may vary depending on the volume of synovial fluid within the myxoid cyst.²¹ In a series of 35 cases of subungual myxoid cysts, none manifested with longitudinal erythronychia. Due to their composition, myxoid cysts can be distinguished easily from solid tumors of the nail unit via transillumination.²² Pain is a much less common with myxoid cysts vs glomus tumors, as the filling of the pseudocyst space with synovial fluid typically is gradual, allowing the surrounding tissue to accommodate and adapt over time.²¹ In equivocal cases, MRI or high-resolution ultrasonography may be used to distinguish myxoid cysts and glomus tumors.⁸ Histopathology shows accumulation of mucin in the dermis with surrounding fibrous stroma.²³

Subungual neuromas are painful benign tumors that develop due to disorganized neural proliferation following disruption to peripheral nerves secondary to trauma or surgery. In 3 case reports, subungual neuromas manifested as painful subungual nodules, with proximal nail plate ridging, or onycholysis.^24-26 Since neuromas have only rarely been described in the subungual region, reports of MRI and ultrasonography findings are unknown. Histopathology is needed to distinguish neuromas from glomus tumors. Histopathology shows an acapsular structure consisting of disorganized spindle-cell proliferation and nerve fibers arranged in a tangle of fascicles within fibrotic tissue.²⁵ On immunochemistry, spindle cells typically are positive for cellular antigen protein S100.²⁶ 

Leiomyomas are benign neoplasms derived from smooth muscle, typically localized to the uterus or gastrointestinal tract, and have been described rarely in the nail unit.^27,28 It is hypothesized that subungual leiomyomas originate from the vascular smooth muscle in the subcutaneous layer of the nail unit.²⁸ Like glomus tumors, leiomyomas of the subungual region often manifest with pain and longitudinal erythronychia.^27-30 Subungual leiomyomas may be distinguished from glomus tumors via advanced imaging techniques, including ultrasonography and MRI. Cutaneous leiomyomas have been described with mild to moderate internal low flow vascularity on Doppler ultrasonography, while glomus tumors typically reveal high internal vascularity.²⁸ Biopsy with histopathology is needed for definitive diagnosis. On histopathology, leiomyomas demonstrate bland-appearing spindle-shaped cells with elongated nuclei arranged in fascicles.²⁷ They typically are positive for α-smooth muscle actin and caldesmon on immunostaining. 

Eccrine spiradenomas are benign adnexal tumors likely of apocrine origin with limited case reports in the literature.^31,32 Clinically, eccrine spiradenomas involving the nail unit may manifest with longitudinal nail splitting of the nail or as a papule on the proximal nail fold, with associated tenderness.^31,32 In a report of a 50-year-old woman with a histopathologically confirmed eccrine spiradenoma manifesting with longitudinal splitting of the nail and pain in the proximal nail fold, the mass appeared hypoechoic on ultrasonography with increased intramass vascularity on Doppler, while MRI showed an intensely enhancing lesion.³¹ These imaging features, combined with a classically manifesting feature of pain, make eccrine spiradenomas difficult to distinguish from glomus tumors; therefore, histopathologic examination can provide a definitive diagnosis, and surgical excision is used for treatment.³¹ On histopathology, these tumors are well circumscribed and composed of both small dark basaloid cells with peripheral compact nuclei and larger cells with central pale nuclei, which may be arranged in tubules.^31,32

The Diagnosis: Glomangiomyoma

The nail unit excision specimen showed collections of cuboidal cells and spindled cells within the corium that were consistent with a diagnosis of a glomangiomyoma, a rare glomus tumor variant (Figure). Glomus tumors are benign neoplasms comprising glomus bodies, which are arteriovenous anastomoses involved in thermoregulation.¹ They develop in areas densely populated by glomus bodies, including the fingers, toes, and subungual areas. Glomus tumors most commonly develop in middle-aged women.² Clinically, they manifest with a characteristic triad of intense pain, point tenderness, and cold sensitivity and may appear as reddish-pink or blue macules under the nail plate and/or longitudinal erythronychia.^2-6 The presence of multiple glomus tumors is associated with neurofibromatosis type 1.⁷ 

Advanced imaging including ultrasonography and magnetic resonance imaging (MRI) may help confirm the diagnosis but may not be cost effective, as excision with histopathology is needed to relieve symptoms and render a definitive diagnosis. Radiography is highly insensitive in identifying bone erosions associated with glomus tumors.⁸ With ultrasonography, glomus tumors appear hypoechoic; with Doppler ultrasonography, they appear hypervascular. With MRI, glomus tumors appear as well-defined nodular lesions with hypointense signal intensity on T1-weighted sequence and hyperintense signal intensity on T2-weighted sequence, with strong enhancement using gadolinium-based contrast.^9,10 On histopathology, a glomus tumor appears as a nodular tumor with sheets of oval-nucleated cells arranged in multicellular layers surrounding blood vessels and are immunoreactive for α-smooth muscle actin, muscle-specific actin, and type IV collagen.^11,12 

There are several glomus tumor variants. The most common is a solid glomus tumor, which predominantly is composed of glomus cells, followed by glomangioma, which mainly is composed of blood vessels. Glomangiomyoma, which mostly is composed of smooth muscle cells, is the rarest variant.¹³ 

While glomus tumors are common in the subungual areas, it is an uncommon location for glomangiomyomas, which have been reported in the nail unit in only 7 prior case reports identified through searches of PubMed and Google Scholar using the terms glomangiomyoma, glomangiomyoma nail, and subungual glomangiomyoma (Table).^13-19 Glomangiomyomas more commonly are described in solid organs, including the stomach, kidney, pancreas, and bladder.¹⁶ The mean age of patients with subungual glomangiomyomas, including our patient, was 40.4 years (range, 3-61 years), with the majority being female (75.0% [6/8]). Most patients presented with fingernail involvement (75.0% [6/8]), nail dystrophy (eg, nail plate thinning, longitudinal grooves, splinter hemorrhages, longitudinal erythronychia)(62.5% [5/8]), and intermittent pain and/or point tenderness in the affected nail (75.0% [6/8]).^13-19 Notably, only our patient had longitudinal erythronychia as a clinical feature, and only one other case described MRI findings, which included a lobulated mass with intense contrast and distal phalanx destruction.¹⁸ One patient was a 3-year-old girl with a family history of generalized multiple glomangiomyomas. Although subungual glomangiomyoma was not confirmed on histopathology, the diagnosis in this patient was presumed based on her family history.¹³ On histopathology, glomangiomyomas are composed of oval-nucleated cells surrounding blood vessels. These oval-nucleated cells then gradually transition to smooth muscle cells.²⁰ 

A myxoid cyst is composed of a pseudocyst, which lacks a cyst lining, and is a result of synovial fluid from the distal interphalangeal joint entering the pseudocyst space.² It typically manifests with a longitudinal groove in the nail plate. A flesh-colored nodule may be appreciated between the cuticle and the distal interphalangeal joint.² The depth of the longitudinal groove may vary depending on the volume of synovial fluid within the myxoid cyst.²¹ In a series of 35 cases of subungual myxoid cysts, none manifested with longitudinal erythronychia. Due to their composition, myxoid cysts can be distinguished easily from solid tumors of the nail unit via transillumination.²² Pain is a much less common with myxoid cysts vs glomus tumors, as the filling of the pseudocyst space with synovial fluid typically is gradual, allowing the surrounding tissue to accommodate and adapt over time.²¹ In equivocal cases, MRI or high-resolution ultrasonography may be used to distinguish myxoid cysts and glomus tumors.⁸ Histopathology shows accumulation of mucin in the dermis with surrounding fibrous stroma.²³

Subungual neuromas are painful benign tumors that develop due to disorganized neural proliferation following disruption to peripheral nerves secondary to trauma or surgery. In 3 case reports, subungual neuromas manifested as painful subungual nodules, with proximal nail plate ridging, or onycholysis.^24-26 Since neuromas have only rarely been described in the subungual region, reports of MRI and ultrasonography findings are unknown. Histopathology is needed to distinguish neuromas from glomus tumors. Histopathology shows an acapsular structure consisting of disorganized spindle-cell proliferation and nerve fibers arranged in a tangle of fascicles within fibrotic tissue.²⁵ On immunochemistry, spindle cells typically are positive for cellular antigen protein S100.²⁶ 

Leiomyomas are benign neoplasms derived from smooth muscle, typically localized to the uterus or gastrointestinal tract, and have been described rarely in the nail unit.^27,28 It is hypothesized that subungual leiomyomas originate from the vascular smooth muscle in the subcutaneous layer of the nail unit.²⁸ Like glomus tumors, leiomyomas of the subungual region often manifest with pain and longitudinal erythronychia.^27-30 Subungual leiomyomas may be distinguished from glomus tumors via advanced imaging techniques, including ultrasonography and MRI. Cutaneous leiomyomas have been described with mild to moderate internal low flow vascularity on Doppler ultrasonography, while glomus tumors typically reveal high internal vascularity.²⁸ Biopsy with histopathology is needed for definitive diagnosis. On histopathology, leiomyomas demonstrate bland-appearing spindle-shaped cells with elongated nuclei arranged in fascicles.²⁷ They typically are positive for α-smooth muscle actin and caldesmon on immunostaining. 

Eccrine spiradenomas are benign adnexal tumors likely of apocrine origin with limited case reports in the literature.^31,32 Clinically, eccrine spiradenomas involving the nail unit may manifest with longitudinal nail splitting of the nail or as a papule on the proximal nail fold, with associated tenderness.^31,32 In a report of a 50-year-old woman with a histopathologically confirmed eccrine spiradenoma manifesting with longitudinal splitting of the nail and pain in the proximal nail fold, the mass appeared hypoechoic on ultrasonography with increased intramass vascularity on Doppler, while MRI showed an intensely enhancing lesion.³¹ These imaging features, combined with a classically manifesting feature of pain, make eccrine spiradenomas difficult to distinguish from glomus tumors; therefore, histopathologic examination can provide a definitive diagnosis, and surgical excision is used for treatment.³¹ On histopathology, these tumors are well circumscribed and composed of both small dark basaloid cells with peripheral compact nuclei and larger cells with central pale nuclei, which may be arranged in tubules.^31,32