AI Applications in Oncology Imaging

📢 New Publication Alert! Proud to share our latest study published in Journal of Clinical Oncology Clinical Cancer Informatics: “Computational Morphological Assessment of Bladder Cancer Tissue Is Prognostic of Recurrence and Overall Survival Following Transurethral Resection” 🔗 https://lnkd.in/eAykZtYc In this study, we developed an interpretable, machine learning-based model that evaluates nuclear pleomorphism, polarity, and N/C ratio from standard H&E-stained slides to predict recurrence and overall survival in bladder cancer patients post-TURBT. 💡 Key Takeaways: 🎯 Developed a feature-based risk model from 430 computational pathology descriptors 🎯 Validated the model on an independent cohort (n=151), achieving up to 0.78 accuracy 🎯 Demonstrated strong alignment with pathologist grading, even in cases of interobserver disagreement 🎯 Offers an interpretable, low-cost, clinically deployable tool with no added tests beyond routine pathology This work underscores the promise of explainable AI in enhancing cancer risk stratification—not replacing pathologists, but augmenting clinical decision-making. Grateful to work with an incredible team including Patrick Leo, Behtash Nezami, Mahmut Akgul, Naoto Tokuyama, Xavier Farré, Vidya Sankar Viswanathan, Gregory McLennan, and more. #BladderCancer #DigitalPathology #MachineLearning #Oncology #ComputationalPathology #AIinMedicine #PrecisionOncology #InterpretableAI #JCOCCI #TeamScience

Hello my Artificial Intelligence enthusiasts! This published study demonstrates the potential of AI to enhance patient care by enabling more precise and effective treatments, reducing adverse side effects. A study published in the Journal of Urology demonstrates that artificial intelligence (AI) software significantly enhances prostate cancer mapping accuracy, outperforming seasoned urologists and radiologists. The AI-assisted approach achieved an 84.7% accuracy rate in defining cancer margins, compared to 67.2% and 75.9% for manually defined margins. This improvement could expand the use of focal therapy in prostate cancer patients, as accurate contouring is critical for successful treatment. The study involved ten experienced clinicians who first manually defined cancer margins using MRI, followed by the application of Avenda Health’s AI-powered software, Unfold AI, on the same cases. The AI not only improved accuracy but also addressed the underestimation of cancer extent, achieving a 72.8% negative margin rate compared to 1.6% manually. The authors highlight the potential of AI to enhance patient care by enabling more precise and effective treatments, reducing adverse side effects. #digitalhealth #AI #MAIC #digitaltransformation #Prostatecancer #diagnostics #cancer #Bayer #artificialintelligence

This is such an impressive development in the Oncology space! Dana Farber's newly developed AI program named OncoNPC, aids in the diagnosis of cancers of unknown primary origin. Such malignancies are hard to trace back to their original source, and this issue impacts 3% to 5% of patients, often limiting treatment options. OncoNPC uses sequencing data from tumor DNA, combined with medical records and genetic testing results from over 36,400 patients, to predict where the cancer originated. This means that the algorithm could potentially more than double the number of patients eligible for approved precision treatments by accurately predicting the origin of about 80% of tumors. Future plans include improving the model by integrating more information, such as pathology reports, and exploring its synergy with other diagnostic techniques. https://lnkd.in/dEGF5xYf

South Korean gave regulatory approval for an 𝗔𝗜-𝗽𝗼𝘄𝗲𝗿𝗲𝗱 𝘀𝗸𝗶𝗻-𝗰𝗵𝗲𝗰𝗸𝗶𝗻𝗴 𝗔𝗣𝗣 𝗳𝗼𝗿 𝘀𝗸𝗶𝗻 𝗰𝗮𝗻𝗰𝗲𝗿 𝗱𝗶𝗮𝗴𝗻𝗼𝘀𝗶𝘀. LifeSemantics secured the first-ever approval for an AI-powered solution for skin cancer detection in the country! The approved solution, canofyMD SCAI, is a cutting-edge mobile app designed to diagnose various types of skin cancer using advanced AI techniques. Here’s the lowdown: ↳ The Technology Is Developed under the Doctor Answer 2.0 initiative, this solution harnesses the power of convolutional neural networks (CNNs) to accurately detect skin cancer from images taken via smartphones. ↳ They Conducted the first clinical trial of its kind in South Korea, using 6,500 images of skin lesions from three local hospitals. The AI was tested on 199 cases, achieving an impressive diagnostic accuracy of 80.9%. They are gearing up for further trials with six local hospitals and aims to roll out the app commercially by the end of this year. They’re also eyeing regulatory approvals in Australia and New Zealand, where skin cancer rates are particularly high. Why it matters ? ↳ Skin cancer cases are rising sharply in South Korea, with a 34% increase in recent years. ↳ Globally, new skin cancer cases top 1.5 million annually, making early detection more crucial than ever. ↳ The skin cancer diagnosis market is booming, projected to grow from $3 billion in 2021 to $5 billion by 2028. This breakthrough not only promises to enhance early detection but also brings hope to millions at risk. Their innovative use of AI could be a game-changer in how we approach skin cancer diagnosis and treatment. Other countries are also innovating in this space. For instance; ↳ The US has approved DermaSensor, an AI-enabled device for primary care that detects skin cancer with up to 96% accuracy. ↳ Australia is working on adapting desktop systems for UV radiation measurement to smartphones, integrating AI to identify potential early indicators of skin cancer. As LifeSemantics continues to push boundaries, it’s exciting to see how AI and technology are reshaping healthcare globally. Follow me for more updates on these revolutionary developments in Healthcare

A new, robust study out of Sweden finds that AI enhances doctor's ability to diagnose breast cancers, and seemingly did so without increasing the rate of false positives (always a worry with breast cancer screening shifts). The question, now, is whether detecting more cancers translates into lives saved and better overall health. I took a look at the study and what it may or may not mean for how mammograms gets read in the future, and why its essential to run these kinds of gold standard trials for AI in healthcare: https://lnkd.in/gfcABHsM

AI based histologic biomarker for prognosis of invasive breast cancer A new AI (Artificial Intelligence) tool may make it possible to spare breast cancer patients unnecessary chemotherapy treatments by using a more precise method of predicting their outcomes, reports a new study. AI evaluations of patient tissues were better at predicting the future course of a patient’s disease than evaluations performed by expert pathologists. The AI tool was able to identify breast cancer patients who are currently classified as high or intermediate risk but who become long-term survivors. That means the duration or intensity of their chemotherapy could be reduced. This is important since chemotherapy is associated with unpleasant and harmful side effects such as nausea, or more rarely, damage to the heart. Currently pathologist evaluate cancerous cells in a patient’s tissue to determine treatment. But patterns of non-cancerous cells are very important in predicting outcomes, the study showed. This is the first study to use AI for comprehensive evaluation of both the cancerous and non-cancerous elements of invasive breast cancer. #ScienceMission #sciencenewshighlights https://lnkd.in/d_mpCmVG