Machine learning tool for diagnosis of cutaneous T-cell lymphoma (CTCL)

This technology is a machine learning tool that improves the diagnosis of cutaneous T-cell lymphoma (CTCL) by integrating the analysis of genetic markers with patient electronic health records.

Unmet Need: Genetic biomarkers and treatment guide for cutaneous T-cell lymphoma

Cutaneous T-cell lymphoma (CTCL) is a rare skin cancer that is often misdiagnosed for other skin disorders. The mechanisms underlying CTCL development are poorly understood, which complicates its diagnosis and treatment. Early detection, diagnosis, and treatment of CTCL are critical in providing care for patients, particularly those with more aggressive forms of the disease. Identifying genetic markers and mutations associated with CTCL could help ensure a faster and more accurate diagnosis, guide treatment decisions, and differentiate between early and more advanced stages of the disease. This would improve patient stratification for both treatment and prognosis.

The Technology: Machine learning and artificial intelligence tool for T-cell lymphoma diagnosis

This technology is a machine learning and artificial intelligence tool that improves the diagnosis of cutaneous T-cell lymphoma (CTCL). Specifically, the technology identifies gene mutations associated with disease survival, prognosis, and severity and uses natural language processing to analyze electronic health records, linking pre-diagnosis criteria with subsequent CTCL diagnosis. By integrating multimodal patient data, this technology improves predictions related to disease development, survival, treatment response, and clinical outcomes. As a result, this approach can aid in timely and accurate CTCL diagnosis and help stratify patients into appropriate treatment groups depending on disease severity and genetic profile.

This technology has been validated using human patient data and electronic health records.

Applications:

• Diagnostic assay for cutaneous T-cell lymphoma (CTCL)
• Diagnostic assay for disease staging and prognosis
• Criteria and assessment tool for cancer treatment
• Research tool for the study of CTCL development, diagnosis, prognosis, and treatment
• Research tool for drug development
• Precision medicine approach and framework for identifying genetic markers in rare diseases
• Clinical tool for survival analysis and prognoses

Advantages:

• Integration of multimodal patient data, medical records, and genomics analyses
• Can be combined with other diagnostic tools and treatment approaches
• Enables differentiation of early and late-stage disease
• Can assist in differentiating CTCL from other skin disorders
• Enables earlier treatment interventions
• Cost-effective and scalable

Lead Inventor:

Larisa Geskin, MD

Related Publications:

• Schrediah CM, DeStephano DM, Pan SS, Wang S, Shen H, Ta CN, Reynolds GB, Fahmy LM, Gordon ER, Adeuyan O, Kwinta BD, Stonesifer CJ, Chan WH, Choi J, Duvic M, Gallardo F, Girardi M, Guitart J, Kim YH, Khodadoust MS, Najidh S, Ni X, Pujol RM, Tensen CP, Vermeer MH, Whittaker S, Tatonetti NP, Chase HS, Pe’er I, Geskin LJ. “Machine learning-based survival analysis reveals prognostic clinical and genetic insights for patients with cutaneous T-cell lymphoma.” Blood. 2023 Nov 2; 142(1): 1715.

• Gordon ER, Trager MH, Ta C, Liu C, Schreidah CM, Adeuyan O, Chase H, Weng C, Geskin LJ. “Supervised machine learning and natural language processing identify early clues of CTCL.” J. Investig. Dermatol. 2024 Aug; 144(8):S34.

• Schreidah CM, DeStephano D, Pan SS< Wang S, Reynolds GB, Fahmy L, Gordon ER, Adeuyan O, Choi J, Duvic M, Gallardo F, Girardi M, Guitart J, Kim Y, Khodadoust M, Najih S, Ni X, Pujol RM, Tensen CP, Vermeer MH, Geskin LJ. “Evaluating time-to-event survival in cutaneous t-cell lymphoma: A clinical-genetic machine learning analysis.” J. Investig. Dermatol. 2024 Aug; 144(8):S149.

Tech Ventures Reference:

• IR CU24366

• Licensing Contact: Cynthia Lang