AsianScientist (May. 31, 2024) – Researchers at Kyushu University Japan have developed a machine-learning model for accurate prognosis of Osteosarcoma, a kind of bone cancer. The model helps improve tumour detection and thereby create a personalized treatment. The study was published in npj Precision Oncology.
Osteosarcoma is usually treated with surgery or chemotherapy, which has improved patient outcomes to a significant level. However, predicting patient prognosis remains a challenge. In traditional methods, the prognosis assessments mainly depend on the rate of necrosis, where a pathologist analyses the part of dead tissue within a tumour. But the reliability of this method is affected by the pathologist’s level of skill or interpretations because different pathologists may interpret the results differently. Consequently, it may not provide an accurate indication of how well a treatment is working.
Keeping this limitation in mind, Dr Kengo Kawaguchi and Dr Kazuki Miyama, co-first authors of the study, along with Dr. Makoto Endo, all from the Department of Orthopaedic Surgery, Graduate School of Medical Sciences, Kyushu University in Japan turned to artificial intelligence (AI) for precisely evaluating the disease. They used a novel approach to predict the prognosis of Osteosarcoma by focusing on the viable tumour cell density after treatment.
In the first phase of the study, they trained a deep-learning model to detect the surviving tumour cells in pathological images. Their AI model showed a great level of proficiency that aligned with the capabilities of expert pathologists. After that, they started analyzing the disease-specific and metastasis-free survival, which are important indicators of patient prognosis. Also, the researchers studied the correlation between AI-estimated viable tumour cell density and prognosis, which revealed promising results.
Patients were divided into groups based on the density of viable tumour cells. Those with high viable tumour cell density had a worse prognosis than those with a lower density of viable tumour cells. Interestingly, the necrosis (cell death) was found to be unrelated to disease-specific survival or metastasis-free survival, which indicates the superiority of AI-based prognosis assessments.
In an article published on Kyushu University’s website, Dr Endo emphasizes the significance of their findings, stating, “In the traditional method, the necrosis rate is calculated as a necrotic area rather than individual cell counts, which is not sufficiently reproducible between assessors and does not adequately reflect the effects of anticancer drugs. We therefore considered using AI to improve the estimation.”
This research has important implications. Using AI in pathology analysis can improve how accurately clinicians detect tumours, decrease differences in opinions between pathologists, and provide quicker prognosis predictions. Also, looking at the density of viable tumour cells, which determines the tumour cell growth after treatment, is a better way to predict how well treatment will work compared to just looking at cell death.
Dr. Endo said, “This new approach has the potential to enhance the accuracy of prognoses for osteosarcoma patients treated with chemotherapy. In the future, we intend to actively apply AI to rare diseases such as Osteosarcoma, which have seen limited advancements in epidemiology, pathogenesis, and etiology. Despite the passage of decades, particularly in treatment strategies, substantial progress remains elusive. By putting AI to the problem, this might finally change”
—
Source: Kyushu University ; Image: National Cancer Institute/Unsplash
The article can be found at: Viable tumor cell density after neoadjuvant chemotherapy assessed using deep learning model reflects the prognosis of Osteosarcoma.
Disclaimer: This article does not necessarily reflect the views of AsianScientist or its staff.
