Dr. Knezevich primarily studies melanoma, a type of skin cancer, along with the techniques used to diagnose and classify skin lesions. He has developed new methods using machine learning to enhance image analysis for more accurate diagnoses. His research addresses key issues like the over-diagnosis of low-risk melanomas and the development of standardized reporting systems that help doctors categorize skin lesions more effectively. His work aims to improve patient care by making diagnoses clearer and more consistent, which is especially beneficial for patients with potentially confusing or challenging skin conditions.
Key findings
In a study on low-risk T1a melanomas, Dr. Knezevich found a 100% survival rate, suggesting many of these cases may not require aggressive treatment.
His new MPATH-Dx version 2.0 schema simplifies lesion classification, which can lead to clearer diagnoses and improved patient care.
Using machine learning techniques, he increased the accuracy of melanoma diagnoses by 6% through better analysis of skin biopsy images.
Dr. Knezevich's Scale-Aware Transformers outperformed five other leading methods, matching the diagnostic accuracy of 187 experienced pathologists.
In a study of nearly 9,000 cases, he found that second opinions from specialized dermatopathologists reduced misclassification rates to 36.7% compared to 52.8% from general pathologists.
Frequently asked questions
Does Dr. Knezevich study melanoma?
Yes, Dr. Knezevich specializes in melanoma and is focused on improving its diagnosis and treatment.
What treatments has Dr. Knezevich researched?
His research focuses on diagnosing melanoma and improving patient outcomes rather than specific treatments.
Is Dr. Knezevich's work relevant for patients with skin lesions?
Yes, his work is highly relevant for patients with skin lesions, particularly those that may be melanoma or benign moles.
How does Dr. Knezevich's research help pathologists?
His research provides pathologists with better tools and methods for diagnosing skin conditions, leading to more accurate results.
Can Dr. Knezevich's findings reduce unnecessary treatments?
Yes, his work aims to clarify which melanomas are truly harmful, potentially reducing unnecessary panic and treatment for patients diagnosed with low-risk lesions.
Publications in plain English
MPATH-Dx version 2.0 schema for melanocytic lesions: A robust tool for standardized diagnostic reporting.
Plain English Researchers have developed a new version of a reporting system called MPATH-Dx for assessing skin lesions, simplifying it from five categories to four. This change aims to make it easier for doctors to agree on diagnoses and improve patient care. The revised system specifically defines criteria for classifying less serious lesions and even identifies certain low-risk melanomas that might not need aggressive treatment.
Who this helps: This helps doctors, patients with skin lesions, and those monitoring melanoma risks.
Melanoma in situ and low-risk pT1a melanoma: Need for new diagnostic terminology.
2025
Clinics in dermatology
Elder DE, Barnhill RL, Eguchi M, Elmore JG, Kerr KF +1 more
Plain English Researchers studied the increasing number of melanoma diagnoses and found that many cases labeled as melanoma might not actually be harmful, leading to unnecessary panic and treatment. They discovered that certain low-risk melanomas, specifically those classified as T1a, had a 100% survival rate and often exhibited characteristics that previously would have classified them simply as benign moles. This matters because changing the way we classify these conditions can reduce the stress and stigma associated with a cancer diagnosis for many patients who are not actually at risk.
Who this helps: This helps patients who receive a melanoma diagnosis but are at very low risk, as well as doctors aiming to provide clearer diagnoses.
VSGD-Net: Virtual Staining Guided Melanocyte Detection on Histopathological Images.
2023
IEEE Winter Conference on Applications of Computer Vision. IEEE Winter Conference on Applications of Computer Vision
Liu K, Li B, Wu W, May C, Chang O +4 more
Plain English This study focused on identifying melanocytes, which are important for diagnosing skin cancer, specifically melanoma. The researchers developed a new method called VSGD-Net that works with standard H&E stained images to virtually enhance the identification of melanocytes, eliminating the need for more complex and costly staining techniques. Their approach significantly improved detection accuracy compared to existing methods, making it a valuable tool for pathologists.
Who this helps: This benefits pathologists and patients by making melanoma diagnosis easier and more reliable.
Segmenting Skin Biopsy Images with Coarse and Sparse Annotations using U-Net.
2022
Journal of digital imaging
Nofallah S, Mokhtari M, Wu W, Mehta S, Knezevich S +5 more
Plain English This research studied how to improve the process of analyzing skin biopsy images, which is crucial for diagnosing melanoma, a type of skin cancer that has been rising in prevalence. The researchers developed a method that uses fewer detailed labels to train a computer program to identify important features in these images effectively. Their approach showed good results in accurately segmenting the images, which can streamline the diagnosis process and potentially lead to better patient outcomes.
Who this helps: This benefits doctors and pathologists who diagnose melanoma and other skin conditions.
Improving the Diagnosis of Skin Biopsies Using Tissue Segmentation.
2022
Diagnostics (Basel, Switzerland)
Nofallah S, Li B, Mokhtari M, Wu W, Knezevich S +4 more
Plain English This study looked at how using machine learning to analyze skin biopsy images can help doctors better diagnose invasive melanoma, a serious skin cancer. By focusing on specific parts of the tissue images, researchers found that the accuracy of diagnoses improved by 6%. This matters because better diagnoses could lead to earlier and more effective treatment for patients with melanoma.
Who this helps: Patients with skin cancer.
Machine learning techniques for mitoses classification.
2021
Computerized medical imaging and graphics : the official journal of the Computerized Medical Imaging Society
Nofallah S, Mehta S, Mercan E, Knezevich S, May CJ +4 more
Plain English This study looked at how well different computer algorithms, known as convolutional neural networks (CNNs), can identify mitosis (the process of cell division) in biopsy images for skin and breast cancer. The researchers found that the ESPNet and DenseNet models performed exceptionally well, with sensitivity rates of 97.6% and 96.8% for melanoma images, and 86.6% and 91.6% for breast cancer images, respectively. This is important because accurate detection of mitosis can help doctors make better diagnoses and treatment decisions for cancer patients.
Who this helps: Patients and doctors involved in cancer diagnosis and treatment.
Classifying Breast Histopathology Images with a Ductal Instance-Oriented Pipeline.
2021
Proceedings of the ... IAPR International Conference on Pattern Recognition. International Conference on Pattern Recognition
Li B, Mercan E, Mehta S, Knezevich S, Arnold CW +3 more
Plain English This study developed a new computer program called the Ductal Instance-Oriented Pipeline (DIOP) that helps classify images of breast tissue to assist in diagnosing cancer. The program identifies individual ducts within the tissue and gathers detailed information, achieving accuracy in classifying types of breast tissue similar to that of expert pathologists. DIOP runs quickly on standard computers, making it practical for use in medical settings.
Who this helps: This benefits doctors and pathologists who analyze breast tissue for cancer diagnosis.
Scale-Aware Transformers for Diagnosing Melanocytic Lesions.
2021
IEEE access : practical innovations, open solutions
Wu W, Mehta S, Nofallah S, Knezevich S, May CJ +3 more
Plain English This study focused on improving the diagnosis of skin lesions related to melanoma, which is often difficult and sometimes inconsistent among doctors. Researchers developed a new computer program that analyzes digital images of skin biopsies and found that it performed better than five other leading methods, matching the accuracy of 187 experienced pathologists. This improvement could lead to more consistent and accurate diagnoses for patients suspected of having melanoma.
Who this helps: Patients at risk for melanoma and their doctors.
Assessment of Second-Opinion Strategies for Diagnoses of Cutaneous Melanocytic Lesions.
2019
JAMA network open
Piepkorn MW, Longton GM, Reisch LM, Elder DE, Pepe MS +10 more
Plain English This study looked at how getting second opinions from experts affects the accuracy of diagnosing skin lesions, specifically those related to melanoma. Out of nearly 9,000 cases examined, almost 43% of pathologists sought second opinions, especially for tricky cases. The findings showed that misclassification rates were lower (about 36.7%) when lesions were reviewed by dermatopathologists trained in these specific conditions, compared to higher misclassification rates (52.8%) among general pathologists without specialized training.
Who this helps: This benefits patients and doctors by providing clearer diagnoses for skin lesions, especially those that may be cancerous.
Influence of variability in assessment of Breslow thickness, mitotic rate and ulceration among US pathologists interpreting invasive melanoma, for the purpose of AJCC staging.
2018
Journal of cutaneous pathology
Taylor L, Hood K, Reisch L, Elmore J, Piepkorn M +4 more
Plain English Researchers studied how different US pathologists assess important factors in melanoma staging, specifically Breslow thickness, mitotic rate, and ulceration in tumor samples. They found that while pathologists generally agreed on the depth of invasion (Breslow thickness), there was substantial disagreement regarding the presence of mitotic figures, which led to misclassification – 15.6% mistakenly identified weak tumors as more aggressive, and 32% failed to recognize tumors that needed a higher stage. This matters because accurate staging is crucial for determining the best treatment options for patients, and improved standardization in future guidelines should reduce these discrepancies.
Who this helps: This helps patients with melanoma and their doctors by ensuring more accurate diagnoses and treatments.
How concerns and experiences with medical malpractice affect dermatopathologists' perceptions of their diagnostic practices when interpreting cutaneous melanocytic lesions.
2016
Journal of the American Academy of Dermatology
Carney PA, Frederick PD, Reisch LM, Knezevich S, Piepkorn MW +7 more
Plain English This study looked at how fears of being sued for malpractice influence the work of dermatopathologists who examine skin lesions. Out of 207 dermatopathologists surveyed, 33% had been involved in a malpractice lawsuit, and those who were older or less qualified were more likely to have been sued. Many dermatopathologists worry that these fears affect their decision-making, as 64% feel confident in their interpretations but still believe malpractice concerns lead them to order unnecessary tests.
Who this helps: This research benefits dermatopathologists and the patients who rely on their assessments.
The self-reported use of immunostains and cytogenetic testing in the diagnosis of melanoma by practicing U.S. pathologists of 10 selected states.
2016
Journal of cutaneous pathology
Zhao G, Lee KC, Kwon G, Frederick PD, Onega TL +5 more
Plain English This study looked at how U.S. pathologists in 10 states use new testing methods, specifically FISH and CGH, to diagnose melanoma, a type of skin cancer. Out of 207 surveyed pathologists, 38% reported using these tests, and those who did tended to be younger, have specialized training, and work at academic institutions. This is important because using advanced testing can improve melanoma diagnosis, allowing for better patient care.
Who this helps: Patients with melanoma benefit from more accurate diagnoses.
Achieving consensus for the histopathologic diagnosis of melanocytic lesions: use of the modified Delphi method.
2016
Journal of cutaneous pathology
Carney PA, Reisch LM, Piepkorn MW, Barnhill RL, Elder DE +4 more
Plain English This study looked at how experienced skin doctors, called dermatopathologists, agree on diagnosing complex skin lesions, specifically melanocytic lesions, which can be benign or cancerous. They evaluated 240 cases and found that, while most of the time they reached agreement (95% of benign cases and 100% of the most serious cases), there were disagreements in diagnosing conditions with moderate and severe atypia. The reasons for these disagreements often involved different opinions on how serious the lesions were and what they observed in the samples.
Who this helps: This helps doctors who diagnose skin conditions, improving accuracy in recognizing skin cancer.
Karavan M, Compton N, Knezevich S, Raugi G, Kodama S +2 more
Plain English This study looked at how teledermatology (TD) could impact the diagnosis of melanoma among US Veterans in the Pacific Northwest. They found that the rate of melanoma was lower among Veterans using TD—15 cases per 100,000 compared to 57 cases per 100,000 for those who saw doctors in person. However, those using TD had more advanced cases of melanoma at the time of diagnosis, indicating that many Veterans were not taking advantage of the TD services available to them, leading to worse outcomes.
Who this helps: This helps Veterans who need skin checks and treatment options for melanoma.
Loss of the p53/p63 regulated desmosomal protein Perp promotes tumorigenesis.
2010
PLoS genetics
Beaudry VG, Jiang D, Dusek RL, Park EJ, Knezevich S +4 more
Plain English This study looked at the role of a protein called Perp, which is important for cell connections in preventing skin cancer, specifically squamous cell carcinoma. The researchers found that when Perp is absent in mice, it leads to increased tumor formation and growth; they observed a significant tumor development in these mice when compared to those with Perp. This loss of Perp not only disrupts cell connections but also triggers inflammation, which further supports cancer development.
Who this helps: This research benefits patients by improving our understanding of cancer progression and potential treatment strategies.
Functional health status of adolescents after the Fontan procedure -- comparison with their siblings.
2009
The Canadian journal of cardiology
Manlhiot C, Knezevich S, Radojewski E, Cullen-Dean G, Williams WG +1 more
Plain English This study looked at the health status of teenagers who have had the Fontan procedure, a surgery for congenital heart defects, and compared them to their siblings. Out of 68 participants, those with siblings reported significantly lower physical health scores, while both groups were similar in social and emotional well-being. This research highlights the important impact that having siblings can have on how Fontan patients perceive their physical health, indicating a need for tailored support for these adolescents.
Who this helps: This helps patients and their families, as well as doctors working with adolescents post-surgery.
DNA sequence of the translocation breakpoints in undifferentiated embryonal sarcoma arising in mesenchymal hamartoma of the liver harboring the t(11;19)(q11;q13.4) translocation.
2007
Genes, chromosomes & cancer
Rajaram V, Knezevich S, Bove KE, Perry A, Pfeifer JD
Plain English This study looked at a specific type of aggressive liver tumor called undifferentiated embryonal sarcoma that can develop within a normally benign tumor in children, known as mesenchymal hamartoma of the liver. The researchers identified the DNA changes associated with this tumor, specifically the t(11;19) translocation, and found that it disrupts a gene related to cancer development, along with a new region linked to liver tumors. Understanding these genetic changes helps doctors better diagnose and treat patients with this rare cancer.
Who this helps: This helps patients with liver tumors and their doctors.
Concurrent translocation of BCL2 and MYC with a single immunoglobulin locus in high-grade B-cell lymphomas.
2005
Leukemia
Knezevich S, Ludkovski O, Salski C, Lestou V, Chhanabhai M +6 more
Plain English This study looked at high-grade B-cell lymphomas to see how often specific genetic changes, known as translocations, occurred together. Researchers found that in a group of 1350 cases, 55 had notable alterations, with 8 showing a complex gene rearrangement involving the MYC and BCL2 genes linked to cancer growth. Understanding these genetic patterns is crucial for better diagnosing and treating aggressive forms of leukemia and lymphoma.
Who this helps: This helps patients with high-grade B-cell lymphomas and their doctors in making more informed treatment decisions.
N-methyl-D-aspartate-induced nitric oxide release: an in vivo microdialysis study.
1993
Neuroscience
Luo D, Knezevich S, Vincent SR
Plain English This study looked at how a substance called N-methyl-D-aspartate (NMDA) causes the release of nitric oxide in the brains of live animals. The researchers found that when they increased the dose of NMDA, the levels of nitric oxide in the cerebellum also increased, showing a clear connection between NMDA activity and nitric oxide release. This is important as it helps us understand how signaling in the brain works, which could lead to better treatments for conditions affecting brain function.
Who this helps: This helps researchers and doctors working to develop therapies for neurological disorders.
Symptomatic scapholunatotriquetral carpal coalition with fusion of the capitatometacarpal joint. Report of a case.
1990
Clinical orthopaedics and related research
Knezevich S, Gottesman M
Plain English This study looks at a rare wrist condition known as carpal coalition, where certain bones in the wrist fuse together. The researchers reported a case involving a 33-year-old man who experienced symptoms from this fusion, making it clear that this specific combination of bone fusions has not been documented before. Understanding these unique cases is important for proper diagnosis and treatment options.
Who this helps: This benefits patients with similar wrist conditions and their doctors.
Streptococcal toxic shocklike syndrome leading to bilateral lower extremity compartment syndrome and renal failure. Report of a case.
1990
Clinical orthopaedics and related research
Knezevich S, Torch M
Plain English This study looked at a rare case of a 13-year-old girl who developed severe leg problems and kidney failure due to an infection called Streptococcal toxic shocklike syndrome. The doctors acted quickly, which helped the girl recover with few lasting effects. This is important because it highlights the urgency of treating compartment syndrome to prevent serious health issues.
Who this helps: This helps doctors and healthcare providers in treating similar cases effectively.
Total joint replacement: both a miracle and a disappointment.
1989
Ohio medicine : journal of the Ohio State Medical Association
Mallory TH, Knezevich S
Plain English Researchers studied total joint replacements, a common treatment for severe arthritis in major joints. They found that while joint implants can last about 10 to 15 years, older and less active patients tend to have better long-term outcomes than younger and more active ones. A promising method using a biocompatible material called hydroxyapatite may help bones grow into implants more effectively, leading to longer-lasting joint replacements.
Who this helps: This benefits patients needing joint replacements, particularly younger, active individuals.
Use of the Greenfield filter to prevent fatal pulmonary embolism associated with total hip and knee arthroplasty.
1989
The Journal of bone and joint surgery. American volume
Vaughn BK, Knezevich S, Lombardi AV, Mallory TH
Plain English The study looked at the use of a Greenfield filter to prevent fatal blood clots in patients undergoing total hip or knee replacements. Out of 42 high-risk patients who received the filter before surgery, none suffered a fatal blood clot, and only one out of 24 patients who received the filter after surgery did. This shows that the Greenfield filter is a safe and effective way to prevent dangerous blood clots in certain patients.
Who this helps: This helps patients at high risk for blood clots during joint surgeries.