Alzheimer’s disease is the primary cause of dementia worldwide, with an increasing morbidity burden that may outstrip diagnosis and management capacity as the population ages. Current methods integrate patient history, neuropsychological testing and MRI to identify likely cases, yet effective practices remain variably applied and lacking in sensitivity and specificity. Here we report an interpretable deep learning strategy that delineates unique Alzheimer’s disease signatures from multimodal inputs of MRI, age, gender, and Mini-Mental State Examination score. Our framework linked a fully convolutional network, which constructs high resolution maps of disease probability from local brain structure to a multilayer perceptron and generates precise, intuitive visualization of individual Alzheimer’s disease risk en route to accurate diagnosis. The model was trained using clinically diagnosed Alzheimer’s disease and cognitively normal subjects from the Alzheimer’s Disease Neuroimaging Initiative (ADNI) dataset (n = 417) and validated on three independent cohorts: the Australian Imaging, Biomarker and Lifestyle Flagship Study of Ageing (AIBL) (n = 382), the Framingham Heart Study (n = 102), and the National Alzheimer’s Coordinating Center (NACC) (n = 582). Performance of the model that used the multimodal inputs was consistent across datasets, with mean area under curve values of 0.996, 0.974, 0.876 and 0.954 for the ADNI study, AIBL, Framingham Heart Study and NACC datasets, respectively. Moreover, our approach exceeded the diagnostic performance of a multi-institutional team of practicing neurologists (n = 11), and high-risk cerebral regions predicted by the model closely tracked post-mortem histopathological findings. This framework provides a clinically adaptable strategy for using routinely available imaging techniques such as MRI to generate nuanced neuroimaging signatures for Alzheimer’s disease diagnosis, as well as a generalizable approach for linking deep learning to pathophysiological processes in human disease.

IMPORTANCE Convergent biological, epidemiological, and clinical data identified urate elevation as a candidate strategy for slowing disability progression in Parkinson disease (PD).OBJECTIVE To determine the safety, tolerability, and urate-elevating capability of the urate precursor inosine in early PD and to assess its suitability and potential design features for a disease-modification trial. DESIGN, SETTING, AND PARTICIPANTSThe Safety of Urate Elevation in PD (SURE-PD) study, a randomized, double-blind, placebo-controlled, dose-ranging trial of inosine, enrolled participants from 2009 to 2011 and followed them for up to 25 months at outpatient visits to 17 credentialed clinical study sites of the Parkinson Study Group across the United States. Seventy-five consenting adults (mean age, 62 years; 55% women) with early PD not yet requiring symptomatic treatment and a serum urate concentration less than 6 mg/dL (the approximate population median) were enrolled.INTERVENTIONS Participants were randomized to 1 of 3 treatment arms: placebo or inosine titrated to produce mild (6.1-7.0 mg/dL) or moderate (7.1-8.0 mg/dL) serum urate elevation using 500-mg capsules taken orally up to 2 capsules 3 times per day. They were followed for up to 24 months (median, 18 months) while receiving the study drug plus 1 washout month. MAIN OUTCOMES AND MEASURESThe prespecified primary outcomes were absence of unacceptable serious adverse events (safety), continued treatment without adverse event requiring dose reduction (tolerability), and elevation of urate assessed serially in serum and once (at 3 months) in cerebrospinal fluid.RESULTS Serious adverse events (17), including infrequent cardiovascular events, occurred at the same or lower rates in the inosine groups relative to placebo. No participant developed gout and 3 receiving inosine developed symptomatic urolithiasis. Treatment was tolerated by 95% of participants at 6 months, and no participant withdrew because of an adverse event. Serum urate rose by 2.3 and 3.0 mg/dL in the 2 inosine groups (P < .001 for each) vs placebo, and cerebrospinal fluid urate level was greater in both inosine groups (P = .006 and <.001, respectively). Secondary analyses demonstrated nonfutility of inosine treatment for slowing disability.CONCLUSIONS AND RELEVANCE Inosine was generally safe, tolerable, and effective in raising serum and cerebrospinal fluid urate levels in early PD. The findings support advancing to more definitive development of inosine as a potential disease-modifying therapy for PD. TRIAL REGISTRATION clinicaltrials.gov Identifier: NCT00833690

Background Declining physical activity commonly occurs in people with Parkinson disease (PD) and contributes to reduced functional capacity and quality of life. Objective The purpose of this study was to explore the preliminary effectiveness, safety, and acceptability of a mobile health (mHealth)–mediated exercise program designed to promote sustained physical activity in people with PD. Design This was a 12-month single-blind (assessor), pilot, comparative-effectiveness, randomized controlled study. Methods An mHealth-mediated exercise program (walking with a pedometer plus engagement in planned exercise supported by a mobile health application) was compared over 1 year with an active control condition (walking with a pedometer and exercise only). There were 51 participants in a community setting with mild-to-moderately severe (Hoehn and Yahr stages 1–3) idiopathic PD. Daily steps and moderate-intensity minutes were measured using a step activity monitor for 1 week at baseline and again at 12 months. Secondary outcomes included the 6-Minute Walk Test, Parkinson Disease Questionnaire 39 mobility domain, safety, acceptability, and adherence. Results Both groups increased daily steps, moderate-intensity minutes, and 6-Minute Walk Test, with no statistically significant between-group differences observed. In the less active subgroup, changes in daily steps and moderate-intensity minutes were clinically meaningful. An improvement in the Parkinson Disease Questionnaire 39 mobility score favored mHealth in the overall comparison and was statistically and clinically meaningful in the less active subgroup. Limitations The limitation of the current study was the small sample size. Conclusions Both groups improved physical activity compared with expected activity decline over 1 year. The addition of the mHealth app to the exercise intervention appeared to differentially benefit the more sedentary participants. Further study in a larger group of people with low activity at baseline is needed.

Background-The Parkinson's Progression Markers Initiative (PPMI) is an ongoing observational, longitudinal cohort study of participants with Parkinson's disease, healthy controls, and carriers of the most common Parkinson's disease-related genetic mutations, which aims to define biomarkers of Parkinson's disease diagnosis and progression. All participants are assessed annually with a battery of motor and non-motor scales, 123-I Ioflupane dopamine transporter (DAT) imaging, and biological variables. We aimed to examine whether non-manifesting carriers of LRRK2 and GBA mutations have prodromal features of Parkinson's disease that correlate with reduced DAT binding.

Institutionalized patients with HD are more motorically, psychiatrically, and behaviorally impaired than their counterparts living at home. However, motor variables alone predicted institutionalization. Treatment strategies that delay the progression of motor dysfunction in HD may postpone the need for institutionalization.