PET/CT deep learning prognosis for treatment decision support in esophageal squamous cell carcinoma

This dataset contains CT, PET, and PET/CT images of patients diagnosed with esophageal squamous cell carcinoma (ESCC). It is released in conjunction with the study: PET/CT deep learning prognosis for treatment decision support in esophageal squamous cell carcinoma. Insights Into Imaging, 2024. DOI: 10.1186/s13244-024-01737-1. Images of ESCC patients who received two different treatment therapies: namely surgery alone (SA) and surgery followed by postoperative adjuvant chemotherapy (SPOCT), are included. Moreover, the dataset also provides the ESCCPro-SA and ESCCPro-SPOCT models and code, proposed in the aforementioned study, to enable future researchers to replicate the models. Specifically, the “ESCC_OA_Data.rar” file consists of two folders, namely “Patients_SA” and “Patients_SPOCT”, which correspond to the data of ESCC patients who underwent each respective therapy. Each image within the dataset follows a specific naming convention. The initial number in the file name represents the anonymized patient identification (ranging from 1 to 646), followed by the type of image (CT, PET, or PET/CT), and the corresponding image slice number. The final number in the file name is specifically used to differentiate between patients treated with SA (numbers 1-2) and patients treated with SPOCT (numbers 3-4). Please cite the reference if using this dataset: PET/CT deep learning prognosis for treatment decision support in esophageal squamous cell carcinoma. Insights Into Imaging, 2024. DOI: 10.1186/s13244-024-01737-1