Computer Image Analysis for Skin Conditions

Overseen By Raja Sivamani

Age: 18+

Sex: Any

Trial Phase: Academic

Sponsor: Integrative Skin Science and Research

Disqualifiers: Prisoners, Unable to consent, others

No Placebo Group All trial participants will receive the active study treatment (no placebo)

Compensation: Varies

Number of Visits: Unknown

150 Participants Needed

Trial Summary

What is the purpose of this trial?

The study is conducted to determine if image-based computer grading can of acne, melasma, rosacea and seborrheic dermatitis correlate well to expert based clinical severity grading.

Do I need to stop my current medications for this trial?

The protocol does not specify if you need to stop your current medications.

What data supports the idea that Computer Image Analysis for Skin Conditions (also known as: No Intervention) is an effective treatment?

The available research shows that computer image analysis for skin conditions is effective because it can predict clinical management decisions accurately. One study found that using images to predict management decisions directly was more accurate than predicting the diagnosis first. This approach also reduced unnecessary procedures by 24.56%. Another study demonstrated that digital imaging for skin cancer diagnosis had almost complete agreement with traditional clinical consultations, showing its promise in teledermatology. Additionally, a facial skin analysis system on a handheld device showed good agreement with dermatologist assessments, supporting its potential use in clinical settings. ^{1 2 3 4 5}

What safety data exists for computer image analysis in skin conditions?

The safety data for computer image analysis in skin conditions can be informed by various sources. The CAPER Registry collects adverse event data from dermatologic procedures, which can help identify safety issues. Safety of dermatology treatments is assessed through clinical trials, registries, and spontaneous reporting, providing a comprehensive understanding of safety profiles. The FDA evaluates safety using data from clinical trials, postmarketing reports, and registries, and continues to develop tools for risk assessment. These sources collectively contribute to understanding the safety of treatments in dermatology, including computer image analysis. ^{6 7 8 9 10}

Is computer image analysis a promising treatment for skin conditions?

Yes, computer image analysis is a promising treatment for skin conditions. It helps doctors diagnose skin issues like melanoma by analyzing images of the skin. This technology can detect changes in skin lesions over time, which is important for early cancer detection. It can also differentiate between benign and malignant lesions, making it a useful tool for doctors, especially those who are not experts in dermatology. ^{11 12 13 14 15}

Research Team

Raja Sivamani

Principal Investigator

Integrative Skin and Research

Eligibility Criteria

This trial is for adults with acne, rosacea, melasma, or seborrheic dermatitis. Participants must be able to give consent and should not have facial markings like piercings or tattoos that could affect the imaging process on the day of facial photography.

Inclusion Criteria

I have acne, rosacea, melasma, or seborrheic dermatitis.

Exclusion Criteria

I am unable to give consent for medical procedures.

Artificial facial markings on day of facial photography (such as piercings and tattoos) that may interfere with imaging in the opinion of the investigator

Prisoners

Treatment Details

Interventions

• No Intervention

Trial Overview The study aims to see if computer-based image grading can accurately assess the severity of skin conditions such as acne, melasma, rosacea, and seborrheic dermatitis compared to expert clinical evaluations.

Participant Groups

Group I: Seborrheic Dermatitis Experimental Treatment 1 Intervention

Grading of seborrheic dermatitis with Seborrheic Dermatitis Area Severity Index score

Group II: Rosacea Experimental Treatment 1 Intervention

Inflammatory lesion count

Group III: Melasma Experimental Treatment 1 Intervention

Pigment intensity and distribution with use of Melasma Area Severity Index

Group IV: Acne vulgaris Experimental Treatment 1 Intervention

Inflammatory and non-inflammatory lesion counts

Who Is Running the Clinical Trial?

Integrative Skin Science and Research

Lead Sponsor

Trials

33

Recruited

2,000+

Codex Labs

Collaborator

Trials

2

Recruited

190+

Findings from Research

The study involved 30 general practitioners and analyzed 981 digital images of skin lesions, finding that teledermatology was feasible and well-received by GPs, with 77.8% reporting no issues with the process.

However, the diagnostic accuracy raised concerns, as 45.3% of biopsies from the highest risk category (category 3) confirmed malignancy, suggesting potential risks of missed skin cancer cases when relying solely on teledermatology for diagnosis.

NCBI (Pubmed)

pubmed.ncbi.nlm.nih.gov

Feasibility and diagnostic accuracy of teledermatology in Swiss primary care: process analysis of a randomized controlled trial. Tandjung, R., Badertscher, N., Kleiner, N., et al.[2015]

A new automated machine learning approach can directly predict clinical management decisions for skin lesions from images, achieving higher accuracy than traditional methods that first predict the diagnosis before inferring management.

This direct prediction method significantly reduces the rate of unnecessary excisions by 24.56%, and when combined with additional criteria for skin lesions, it further improves accuracy in management predictions.

NCBI (Pubmed)

pubmed.ncbi.nlm.nih.gov

Predicting the clinical management of skin lesions using deep learning. Abhishek, K., Kawahara, J., Hamarneh, G.[2021]

A study involving 12 patients and 13 skin lesions demonstrated that digital image consultations for teledermatology can achieve almost complete agreement in diagnosis and biopsy recommendations compared to traditional clinic-based consultations.

The findings suggest that using inexpensive, off-the-shelf equipment for digital imaging in teledermatology is a promising approach, as it allows dermatologists to accurately assess skin lesions remotely.

NCBI (Pubmed)

pubmed.ncbi.nlm.nih.gov

A pilot trial of digital imaging in skin cancer. Whited, JD., Mills, BJ., Hall, RP., et al.[2017]

A review of 39 studies on computer-aided diagnosis for nonmelanoma skin cancer (NMSC) found that automated detection models achieved high diagnostic accuracy, with results ranging from 72% to 100%, indicating their potential effectiveness compared to human dermatologists.

Despite the promising accuracy of these image-based classifiers, many studies had significant methodological limitations, highlighting the need for more robust clinical trials to evaluate their practical application in real-time diagnosis of NMSC.

NCBI (Pubmed)

pubmed.ncbi.nlm.nih.gov

Automated detection of nonmelanoma skin cancer using digital images: a systematic review. Marka, A., Carter, JB., Toto, E., et al.[2020]

A handheld skin analysis software was validated against assessments from a board-certified dermatologist, showing a good agreement rate of 69% across various skin characteristics in participants of different ages and skin types.

The software demonstrated particularly high agreement for assessing erythema (83.7%) and wrinkles (81.6%), indicating its potential as a reliable tool for objective skin analysis in clinical settings.

NCBI (Pubmed)

pubmed.ncbi.nlm.nih.gov

Initial validation of a new device for facial skin analysis. Cook, MK., Kaszycki, MA., Richardson, I., et al.[2022]

Dermatologists face challenges in assessing drug safety due to the vast number of skin conditions (over 3000 diagnoses), necessitating the use of diverse data sources including clinical trials and real-world evidence.

New therapies like dupilumab for atopic dermatitis and Janus-kinase inhibitors for alopecia areata are supported by evolving high-level evidence, highlighting the importance of both trial data and ongoing safety monitoring in clinical practice.

NCBI (Pubmed)

pubmed.ncbi.nlm.nih.gov

How is safety of dermatology drugs assessed: trials, registries, and spontaneous reporting. Asfour, L., Yiu, ZZN., Warren, RB.[2021]

Current strategies for monitoring adverse events (AEs) in dermatologic surgery are inconsistent, with morbidity and mortality conferences being the most common reporting method, but more sensitive approaches like retrospective medical record reviews and anonymous electronic systems exist.

There is no standard for AE reporting in dermatologic surgery, highlighting the need for pilot studies to establish effective measures that can enhance AE detection and improve patient safety outcomes.

NCBI (Pubmed)

pubmed.ncbi.nlm.nih.gov

Detecting adverse events in dermatologic surgery. Pinney, D., Pearce, DJ., Feldman, SR.[2019]

The FDA continuously evaluates the safety of marketed pharmaceuticals using data from clinical trials, postmarketing reports, and epidemiological studies, ensuring ongoing monitoring throughout a product's life cycle.

For certain drugs, especially in dermatology, new risks may emerge after approval, prompting the FDA to explore advanced tools like pharmacogenomic biomarkers and large medical databases to better identify and manage these safety concerns.

NCBI (Pubmed)

pubmed.ncbi.nlm.nih.gov

Finding, evaluating, and managing drug-related risks: approaches taken by the US Food and Drug Administration (FDA). Weaver, J., Grenade, LL., Kwon, H., et al.[2009]

The CAPER Registry is a national safety reporting program that collects data on adverse events from dermatologic procedures, aiming to enhance safety monitoring in the field.

By identifying gaps in resources, processes, and education related to adverse events, CAPER seeks to improve clinical practices and prevent future incidents, benefiting practitioners, patients, and regulatory bodies.

NCBI (Pubmed)

pubmed.ncbi.nlm.nih.gov

The Cutaneous Procedures Adverse Events Reporting (CAPER) Registry. Kang, BY., Ibrahim, SA., Poon, E., et al.[2022]

Cutaneous adverse drug reactions (CADRs) are common and challenging to predict, but using nonhuman primates (NHPs) in toxicity studies may improve the predictability of these reactions if skin findings are thoroughly analyzed.

A systematic approach to documenting and evaluating skin findings in NHP studies can help distinguish between drug-related and unrelated skin issues, enhancing the relevance of these findings for patient safety in clinical trials.

NCBI (Pubmed)

pubmed.ncbi.nlm.nih.gov

Recommended Diagnostic Approach to Documenting and Reporting Skin Findings of Nonhuman Primates from Regulatory Toxicity Studies. Mecklenburg, L., Romeike, A.[2018]

The developed mathematical algorithms for analyzing digital photographs of skin lesions showed high sensitivity in detecting lesions as small as 2 mm, which is crucial for early skin cancer detection.

In a pilot study with 17 patients, the software helped clinicians identify new or enlarged lesions in 3 to 11 additional patients compared to unregistered images, indicating its potential to enhance clinical assessments of skin changes.

NCBI (Pubmed)

pubmed.ncbi.nlm.nih.gov

Development and Narrow Validation of Computer Vision Approach to Facilitate Assessment of Change in Pigmented Cutaneous Lesions. Maguire, WF., Haley, PH., Dietz, CM., et al.[2023]

A new online system for analyzing dermoscopy images allows engineers to securely register their algorithms and enables users to customize their image analysis, promoting collaboration among researchers to improve melanoma diagnosis.

This platform aims to enhance the accuracy of pigmented skin lesion classification while supporting mass screening efforts, although challenges like standardization of image acquisition and copyright issues need to be addressed.

NCBI (Pubmed)

pubmed.ncbi.nlm.nih.gov

An open Internet platform to distributed image processing applied to dermoscopy. Guillod, J., Schmid-Saugeon, P., Décaillet, F., et al.[2007]

A study using computer-aided image analysis identified 23 significant parameters that can differentiate between pigmented skin lesions like nevus, lentigo, and seborrheic keratosis, enhancing diagnostic accuracy without the need for biopsies.

The analysis revealed that nevus is the darkest with the highest texture contrast, while seborrheic keratosis is larger and spikier, indicating that software can detect subtle differences in skin lesions that may be missed by the human eye.

NCBI (Pubmed)

pubmed.ncbi.nlm.nih.gov

Differentiation of benign pigmented skin lesions with the aid of computer image analysis: a novel approach. Choi, JW., Park, YW., Byun, SY., et al.[2021]

The study utilized advanced digital image analysis techniques on 20 color slides of cutaneous melanoma to objectively evaluate skin lesions, aiming to identify details that are not visible to the human eye.

The findings suggest that this method could enhance the accuracy of melanoma diagnosis and improve clinicopathologic correlations, although further technical improvements and clinical trials are necessary to validate its sensitivity and specificity.

NCBI (Pubmed)

pubmed.ncbi.nlm.nih.gov

A possible new tool for clinical diagnosis of melanoma: the computer. Cascinelli, N., Ferrario, M., Tonelli, T., et al.[2019]

The computerized image analysis system demonstrated high diagnostic efficiency (0.98) for distinguishing between benign and malignant pigmented lesions, based on a training set of 169 lesions from 165 patients.

With a threshold of two positive indicators, the system showed a negative predictive value of 0.95, indicating it is highly reliable in ruling out melanoma, making it a valuable tool for non-expert clinicians in diagnosing skin lesions.

NCBI (Pubmed)

pubmed.ncbi.nlm.nih.gov

Results obtained by using a computerized image analysis system designed as an aid to diagnosis of cutaneous melanoma. Cascinelli, N., Ferrario, M., Bufalino, R., et al.[2019]

References

1. England pubmed.ncbi.nlm.nih.gov

Feasibility and diagnostic accuracy of teledermatology in Swiss primary care: process analysis of a randomized controlled trial. [2015]

2. England pubmed.ncbi.nlm.nih.gov

Predicting the clinical management of skin lesions using deep learning. [2021]

3. England pubmed.ncbi.nlm.nih.gov

A pilot trial of digital imaging in skin cancer. [2017]

4. England pubmed.ncbi.nlm.nih.gov

Automated detection of nonmelanoma skin cancer using digital images: a systematic review. [2020]

5. England pubmed.ncbi.nlm.nih.gov

Initial validation of a new device for facial skin analysis. [2022]

6. England pubmed.ncbi.nlm.nih.gov

How is safety of dermatology drugs assessed: trials, registries, and spontaneous reporting. [2021]

7. United States pubmed.ncbi.nlm.nih.gov

Detecting adverse events in dermatologic surgery. [2019]

8. United States pubmed.ncbi.nlm.nih.gov

Finding, evaluating, and managing drug-related risks: approaches taken by the US Food and Drug Administration (FDA). [2009]

9. Germany pubmed.ncbi.nlm.nih.gov

The Cutaneous Procedures Adverse Events Reporting (CAPER) Registry. [2022]

10. United States pubmed.ncbi.nlm.nih.gov

Recommended Diagnostic Approach to Documenting and Reporting Skin Findings of Nonhuman Primates from Regulatory Toxicity Studies. [2018]

11. Netherlands pubmed.ncbi.nlm.nih.gov

Development and Narrow Validation of Computer Vision Approach to Facilitate Assessment of Change in Pigmented Cutaneous Lesions. [2023]

12. Netherlands pubmed.ncbi.nlm.nih.gov

An open Internet platform to distributed image processing applied to dermoscopy. [2007]

13. Korea (South) pubmed.ncbi.nlm.nih.gov

Differentiation of benign pigmented skin lesions with the aid of computer image analysis: a novel approach. [2021]

14. United States pubmed.ncbi.nlm.nih.gov

A possible new tool for clinical diagnosis of melanoma: the computer. [2019]

15. England pubmed.ncbi.nlm.nih.gov

Results obtained by using a computerized image analysis system designed as an aid to diagnosis of cutaneous melanoma. [2019]