Alzheimer resemblance atrophy index, BrainAGE, and normal pressure hydrocephalus score in the prediction of subtle cognitive decline: added value compared to existing MR imaging markers

MRI-Based Prediction of Clinical Improvement after Ventricular Shunt Placement for Normal Pressure Hydrocephalus: Development and Evaluation of an Integrated Multisequence Machine Learning Algorithm

BACKGROUND AND PURPOSE

Symptoms of normal pressure hydrocephalus (NPH) are sometimes refractory to shunt placement, with limited ability to predict improvement for individual patients. We evaluated an MRI-based artificial intelligence method to predict postshunt NPH symptom improvement.

MATERIALS AND METHODS

Patients with NPH who underwent MRI before shunt placement at a single center (2014-2021) were identified. Twelve-month postshunt improvement in mRS, incontinence, gait, and cognition were retrospectively abstracted from clinical documentation. 3D deep residual neural networks were built on skull-stripped T2-weighted and FLAIR images. Predictions based on both sequences were fused by additional network layers. Patients from 2014-2019 were used for parameter optimization, while those from 2020-2021 were used for testing. Models were validated on an external validation data set from a second institution (n = 33).

RESULTS

Of 249 patients, n = 201 and n = 185 were included in the T2-based and FLAIR-based models according to imaging availability. The combination of T2-weighted and FLAIR sequences offered the best performance in mRS and gait improvement predictions relative to models trained on imaging acquired by using only 1 sequence, with area under the receiver operating characteristic (AUROC) values of 0.7395 [0.5765-0.9024] for mRS and 0.8816 [0.8030-0.9602] for gait. For urinary incontinence and cognition, combined model performances on predicting outcomes were similar to FLAIR-only performance, with AUROC values of 0.7874 [0.6845-0.8903] and 0.7230 [0.5600-0.8859].

CONCLUSIONS

Application of a combined algorithm by using both T2-weighted and FLAIR sequences offered the best image-based prediction of postshunt symptom improvement, particularly for gait and overall function in terms of mRS.

Ventricular CSF proteomic profiles and predictors of surgical treatment outcome in chronic hydrocephalus

BACKGROUND

By applying an unbiased proteomic approach, we aimed to search for cerebrospinal fluid (CSF) protein biomarkers distinguishing between obstructive and communicating hydrocephalus in order to improve appropriate surgical selection for endoscopic third ventriculostomy vs. shunt implants. Our second study purpose was to look for potential CSF biomarkers distinguishing between patients with adult chronic hydrocephalus benefitting from surgery (responders) vs. those who did not (non-responders).

METHODS

Ventricular CSF samples were collected from 62 patients with communicating hydrocephalus and 28 patients with obstructive hydrocephalus. CSF was collected in relation to the patients' surgical treatment. As a control group, CSF was collected from ten patients with unruptured aneurysm undergoing preventive surgery (vascular clipping).

RESULTS

Mass spectrometry-based proteomic analysis of the samples identified 1251 unique proteins. No proteins differed significantly between the communicating hydrocephalus group and the obstructive hydrocephalus group. Four proteins were found to be significantly less abundant in CSF from communicating hydrocephalus patients compared to control subjects. A PCA plot revealed similar proteomic CSF profiles of obstructive and communicating hydrocephalus and control samples. For obstructive hydrocephalus, ten proteins were found to predict responders from non-responders.

CONCLUSION

Here, we show that the proteomic profile of ventricular CSF from patients with hydrocephalus differs slightly from control subjects. Furthermore, we find ten predictors of response to surgical outcome (endoscopic third ventriculostomy or ventriculo-peritoneal shunt) in patients with obstructive hydrocephalus.

Automatic MRI volumetry in asymptomatic cases at risk for normal pressure hydrocephalus

The occurrence of significant Alzheimer's disease (AD) pathology was described in approximately 30% of normal pressure hydrocephalus (NPH) cases, leading to the distinction between neurodegenerative and idiopathic forms of this disorder. Whether or not there is a specific MRI signature of NPH remains a matter of debate. The present study focuses on asymptomatic cases at risk for NPH as defined with automatic machine learning tools and combines automatic MRI assessment of cortical and white matter volumetry, risk of AD (AD-RAI), and brain age gap estimation (BrainAge). Our hypothesis was that brain aging and AD process-independent volumetric changes occur in asymptomatic NPH-positive cases. We explored the volumetric changes in normal aging-sensitive (entorhinal cortex and parahippocampal gyrus/PHG) and AD-signature areas (hippocampus), four control cortical areas (frontal, parietal, occipital, and temporal), and cerebral and cerebellar white matter in 30 asymptomatic cases at risk for NPH (NPH probability >30) compared to 30 NPH-negative cases (NPH probability <5) with preserved cognition. In univariate regression models, NPH positivity was associated with decreased volumes in the hippocampus, parahippocampal gyrus (PHG), and entorhinal cortex bilaterally. The strongest negative association was found in the left hippocampus that persisted when adjusting for AD-RAI and Brain Age values. A combined model including the three parameters explained 36.5% of the variance, left hippocampal volumes, and BrainAge values, which remained independent predictors of the NPH status. Bilateral PHG and entorhinal cortex volumes were negatively associated with NPH-positive status in univariate models but this relationship did not persist when adjusting for BrainAge, the latter remaining the only predictor of the NPH status. We also found a negative association between bilateral cerebral and cerebellar white matter volumes and NPH status that persisted after controlling for AD-RAI or Brain Age values, explaining between 50 and 65% of its variance. These observations support the idea that in cases at risk for NPH, as defined by support vector machine assessment of NPH-related MRI markers, brain aging-related and brain aging and AD-independent volumetric changes coexist. The latter concerns volume loss in restricted hippocampal and white matter areas that could be considered as the MRI signature of idiopathic forms of NPH.

Neuroimaging-Based Brain Age Estimation: A Promising Personalized Biomarker in Neuropsychiatry

It is now possible to estimate an individual's brain age via brain scans and machine-learning models. This validated technique has opened up new avenues for addressing clinical questions in neurology, and, in this review, we summarize the many clinical applications of brain-age estimation in neuropsychiatry and general populations. We first provide an introduction to typical neuroimaging modalities, feature extraction methods, and machine-learning models that have been used to develop a brain-age estimation framework. We then focus on the significant findings of the brain-age estimation technique in the field of neuropsychiatry as well as the usefulness of the technique for addressing clinical questions in neuropsychiatry. These applications may contribute to more timely and targeted neuropsychiatric therapies. Last, we discuss the practical problems and challenges described in the literature and suggest some future research directions.