In the past decade the management of differentiated thyroid cancer (DTC) underwent a paradigm shift towards the use of risk-stratification with the goal of maximizing benefit and minimizing morbidity of radioiodine (¹³¹I) therapy. ¹³¹I therapy is guided by information derived from surgical histopathology, molecular markers, postoperative diagnostic radioiodine scintigraphy and thyroglobulin (Tg) levels. ¹³¹I is used for diagnostic imaging and therapy of DTC based on physiologic sodium-iodine symporter expression in normal and neoplastic thyroid tissue. We summarize the essential information at the core of multidisciplinary DTC management, which emphasizes individualization of ¹³¹I therapy according to the patient’s risk for tumor recurrence.

In recent years, there has been a proliferation in the development of targeted radionuclide cancer therapy. It is now possible to use baseline clinical and imaging assessments to determine the most effective therapy and to tailor this therapy during the course of treatment based on radiation dosimetry and tumor response. Although this personalized approach to medicine has the advantage of maximizing therapeutic effect while limiting toxicity, it can be challenging to implement and expensive. Further, in order to use targeted radionuclide therapy effectively, there is a need for multidisciplinary awareness, education, and collaboration across the scientific, industrial, and medical communities. Even more important, there is a growing understanding that combining radiopharmaceuticals with conventional treatment such as chemotherapy and external beam radiotherapy may limit patient morbidity while improving survival. Developments in radiopharmaceuticals as biomarkers capable of predicting therapeutic response and targeting disease are playing a central role in medical research. Adoption of a practical approach to manufacturing and delivering radiopharmaceuticals, assessing patient eligibility, optimizing post-therapy follow-up, and addressing reimbursement issues will be essential for their success.

This article is the second part in our machine learning series. Part 1 provided a general overview of machine learning in nuclear medicine. Part 2 focuses on neural networks. We start with an example illustrating how neural networks work and a discussion of potential applications. Recognizing there is a spectrum of applications, we focus on recent publications in the areas of image reconstruction, low-dose PET, disease detection and models used for diagnosis and outcome prediction. Finally, since the way machine learning algorithms are reported in the literature is extremely variable, we conclude with a call to arms regarding the need for standardized reporting of design and outcome metrics and we propose a basic checklist our community might follow going forward.

In patients with metastatic melanoma and KIT amplifications and/or mutations, therapy with imatinib mesylate may prolong survival. 18F-labeled 2-fluoro-2-deoxy-D-glucose (18F-FDG) PET/CT may be used to assess metabolic response. We investigated associations of metabolic response, mutational status, progression-free survival and overall survival in this population.

Purpose: To evaluate random forests (RFs) as a supervised machine learning algorithm to classify amyloid brain PET as positive or negative for amyloid deposition and identify key regions of interest for stratification. Methods: The data set included 57 baseline ¹⁸F-florbetapir (Amyvid; Lilly, Indianapolis, IN) brain PET scans in participants with severe white matter disease, presenting with either transient ischemic attack/lacunar stroke or mild cognitive impairment from early Alzheimer disease, enrolled in a multicenter prospective observational trial. Scans were processed using the MINC toolkit to generate SUV ratios, normalized to cerebellar gray matter, and clinically read by 2 nuclear medicine physicians with interpretation based on consensus (35 negative, 22 positive). SUV ratio data and clinical reads were used for supervised training of an RF classifier programmed in MATLAB. Results: A 10,000-tree RF, each tree using 15 randomly selected cases and 20 randomly selected features (SUV ratio per region of interest), with 37 cases for training and 20 cases for testing, had sensitivity = 86% (95% confidence interval [CI], 42%–100%), specificity = 92% (CI, 64%–100%), and classification accuracy = 90% (CI, 68%–99%). The most common features at the root node (key regions for stratification) were (1) left posterior cingulate (1039 trees), (2) left middle frontal gyrus (1038 trees), (3) left precuneus (857 trees), (4) right anterior cingulate gyrus (655 trees), and (5) right posterior cingulate (588 trees). Conclusions: Random forests can classify brain PET as positive or negative for amyloid deposition and suggest key clinically relevant, regional features for classification.

Purpose The aim of this study was to evaluate random forests (RFs) to identify ROIs on ¹⁸F-florbetapir and ¹⁸F-FDG PET associated with Montreal Cognitive Assessment (MoCA) score. Materials and Methods Fifty-seven subjects with significant white matter disease presenting with either transient ischemic attack/lacunar stroke or mild cognitive impairment from early Alzheimer disease, enrolled in a multicenter prospective observational trial, had MoCA and ¹⁸F-florbetapir PET; 55 had ¹⁸F-FDG PET. Scans were processed using the MINC toolkit to generate SUV ratios, normalized to cerebellar gray matter (¹⁸F-florbetapir PET), or pons (¹⁸F-FDG PET). SUV ratio data and MoCA score were used for supervised training of RFs programmed in MATLAB. Results ¹⁸F-Florbetapir PETs were randomly divided into 40 training and 17 testing scans; 100 RFs of 1000 trees, constructed from a random subset of 16 training scans and 20 ROIs, identified ROIs associated with MoCA score: right posterior cingulate gyrus, right anterior cingulate gyrus, left precuneus, left posterior cingulate gyrus, and right precuneus. Amyloid increased with decreasing MoCA score. ¹⁸F-FDG PETs were randomly divided into 40 training and 15 testing scans; 100 RFs of 1000 trees, each tree constructed from a random subset of 16 training scans and 20 ROIs, identified ROIs associated with MoCA score: left fusiform gyrus, left precuneus, left posterior cingulate gyrus, right precuneus, and left middle orbitofrontal gyrus. ¹⁸F-FDG decreased with decreasing MoCA score. Conclusions Random forests help pinpoint clinically relevant ROIs associated with MoCA score; amyloid increased and ¹⁸F-FDG decreased with decreasing MoCA score, most significantly in the posterior cingulate gyrus.