The journal of Artificial Intelligence In Oncology (AIO), established in 2017, serves as a new forum for the publication of academic and industry-related papers in the field of AI and its application in medical oncology and basic cancer research. The journal is particularly interested in publishing conceptualization and complementing of AI algorithms and reasoning that will promote the application of AI in facilitating the diagnosis and predicting prognosis of malignancies.
The inherent complexity of human malignancies call for the development of new cutting-edge technologies, concepts and methods that will eventually be applied in diagnosis and treatment of cancer patients. The journal of Artificial Intelligence in Oncology (AIO) is established to provide the unmet academic requirements in publication, evaluation and quality control of AI's applications in the field of medical oncology.
Aims and Scope
Artificial Intelligence in Oncology covers a broad area related with the application of artificial intelligence in cancer management. Particular attention is given to manuscripts pertaining to:
- AI-based cancer diagnosis;
- AI-based design of treatment plans;
- Data-mining and algorithms of cancer epidemiology;
- AI-based analysis of genomics;
- Data mining of genomics;
- Agent-based systems and application in medical oncology;
- AI-based evaluation of bio-markers;
- Conceptualization of AI methodologies in oncology;
- AI-based diagnosis and case studies;
- Machine learning and cancer patient care;
- Automated patient-machine interface;
- Automated positron emission tomography/ computed tomography (PET/CT) contouring;
- Automated evidence-based chemo-/radio-dosage suggestions;
- Automated radiation prescription;
- Automated PET/CT/MRI image analysis
The inaugural issue of the Artificial Intelligence in Oncology
Editorial Insights on Artificial Intelligence in Oncology
Dr. Thrun and his team developed a AI system that can classify skin lesions using a single CNN, trained end-to-end from images directly, using only pixels and disease labels as inputs. They demonstrated that the CNN they developed has achieved performance comparable with dermatologists in the field. Their discovery will make a smartphone-based early diagnosis system for skin lesions that have similar level of competence as a trained dermatologists. Their work is published in the journal Nature.
Dr. Zhang and his team developed an AI system in classfying age-related macular degeneration and diabetic macular edema using transfer learning techniques. This technique may help early diagnosis and treatment of different types of treatable diease including age-related macular degeneration, diabetic macular edema and also pediatric pneumonia using chest X-ray images as demonstrated in the paper. The work is published in the journal Cell.
Dr.Bogani and his team developed artificial neuronal network (ANN) to study whether the pretreatment human papillomavirus (HPV) genotype might predict the risk of having cervical dysplasia persistence/recurrence. Thier ANN system identified key genotype features that will predict cervical dysplasia persistence/recurrence. These genotypes include HPV-16, HPV-59, HPV-52, HPV-18. These genotypes were identified correlated with an increased risk of cervical dysplasia persistence/recurrence. See the full story in the journal European Journal of Cancer Prevention.
Dr. Zhavoronkov and his team identified new cancer biomarker associated with the mammalian embryonic-fetal transition (EFT) using deep neural network (DNN) ensembles trained on analyze transcriptomic data. Their DNN system identified COX7A1 gene as a potential EFT marker which may potentially be used as a target for controlling the embryonic-fetal transition. Their work is publised in the journal Oncotarget.
Dr. Jeroen van der Laak and the team developed automated deep learning algorithms in detecting metastases in hematoxylin and eosin–stained tissue sections of lymph nodes from women with breast cancer. The results generated by AI were compared with pathologists’ diagnoses in clincial setting and they found for the whole-slide image classification task, AI algorithms significantly out-performed pathologists. Their work is publised in the journal JAMA.
Dr. Wong and his team used deep learning system (DLS) in multiethnic cohorts of patients with diabetes to study the sensitivity and specificity of AI in identifying diabetic retinopathy and related eye diseases. They found as quoted: 'the DLS had a sensitivity of 90.5% and specificity of 91.6% for detecting referable diabetic retinopathy; 100% sensitivity and 91.1% specificity for vision-threatening diabetic retinopathy; 96.4% sensitivity and 87.2% specificity for possible glaucoma; and 93.2% sensitivity and 88.7% specificity for age-related macular degeneration, compared with professional graders.' See full story in JAMA.