A natural language processing pipeline for pairing measurements uniquely across free-text CT reports.

OBJECTIVE: To standardize and objectivize treatment response assessment in oncology, guidelines have been proposed that are driven by radiological measurements, which are typically communicated in free-text reports defying automated processing. We study through inter-annotator agreement and natural language processing (NLP) algorithm development the task of pairing measurements that quantify the same finding across consecutive radiology reports, such that each measurement is paired with at most one other ("partial uniqueness"). METHODS AND MATERIALS: Ground truth is created based on 283 abdomen and 311 chest CT reports of 50 patients each. A pre-processing engine segments reports and extracts measurements. Thirteen features are developed based on volumetric similarity between measurements, semantic similarity between their respective narrative contexts and structural properties of their report positions. A Random Forest classifier (RF) integrates all features. A "mutual best match" (MBM) post-processor ensures partial uniqueness. RESULTS: In an end-to-end evaluation, RF has precision 0.841, recall 0.807, F-measure 0.824 and AUC 0.971; with MBM, which performs above chance level (P<0.001), it has precision 0.899, recall 0.776, F-measure 0.833 and AUC 0.935. RF (RF+MBM) has error-free performance on 52.7% (57.4%) of report pairs. DISCUSSION: Inter-annotator agreement of three domain specialists with the ground truth (κ>0.960) indicates that the task is well defined. Domain properties and inter-section differences are discussed to explain superior performance in abdomen. Enforcing partial uniqueness has mixed but minor effects on performance. CONCLUSION: A combined machine learning-filtering approach is proposed for pairing measurements, which can support prospective (supporting treatment response assessment) and retrospective purposes (data mining).

Automatically pairing measured findings across narrative abdomen CT reports.

Radiological measurements are one of the key variables in widely adopted guidelines (WHO, RECIST) that standardize and objectivize response assessment in oncology care. Measurements are typically described in free-text, narrative radiology reports. We present a natural language processing pipeline that extracts measurements from radiology reports and pairs them with extracted measurements from prior reports of the same clinical finding, e.g., lymph node or mass. A ground truth was created by manually pairing measurements in the abdomen CT reports of 50 patients. A Random Forest classifier trained on 15 features achieved superior results in an end-to-end evaluation of the pipeline on the extraction and pairing task: precision 0.910, recall 0.878, F-measure 0.894, AUC 0.988. Representing the narrative content in terms of UMLS concepts did not improve results. Applications of the proposed technology include data mining, advanced search and workflow support for healthcare professionals managing radiological measurements.

Dialectical behavior therapy for borderline personality disorder.

Since the introduction of Linehan's treatment manuals in 1993, dialectical behavior therapy (DBT) has been widely disseminated throughout multiple therapeutic settings and applied to a variety of diagnoses. The enthusiasm with which it was embraced by clinicians early on led some to question whether DBT's popularity was outstripping its empirical foundation. Most of the specific concerns raised regarding DBT's early empirical base have been meaningfully addressed in subsequent randomized controlled trials. This review provides a brief introduction to DBT, followed by a critical appraisal of empirical support for the treatment and a discussion of current research trends.

Olanzapine and fluoxetine administration and coadministration increase rat hippocampal pregnenolone, allopregnanolone and peripheral deoxycorticosterone: implications for therapeutic actions.

Olanzapine and fluoxetine elevate the GABAergic neuroactive steroid allopregnanolone to physiologically relevant concentrations in rodent cerebral cortex. It is unknown if these agents also alter pregnenolone or deoxycorticosterone. Since olanzapine and fluoxetine in combination have clinical utility and may demonstrate synergistic effects, we investigated neuroactive steroid alterations following olanzapine, fluoxetine or coadministration. Male rats received IP vehicle, olanzapine, fluoxetine or the combination of both agents in higher-dose (0, 10, 20 or 10/20 mg/kg, respectively) and lower-dose (0, 5, 10 or 5/10 mg/kg, respectively) experiments. Pregnenolone and allopregnanolone levels in hippocampus were determined by gas chromatography/mass spectrometry. Peripheral deoxycorticosterone and other steroid levels were determined by radioimmunoassay. Olanzapine, fluoxetine or the combination increased hippocampal pregnenolone and serum deoxycorticosterone in both higher- and lower-dose experiments, and elevated hippocampal allopregnanolone in higher-dose conditions. No synergistic effects on pregnenolone or allopregnanolone were observed following olanzapine and fluoxetine coadministration compared to either compound alone. Pregnenolone and its sulfate enhance learning and memory in rodent models, and therefore pregnenolone elevations may be relevant to cognitive changes in psychotic and affective disorders. Since pregnenolone decreases have been linked to depression, it is possible that olanzapine- and fluoxetine-induced pregnenolone elevations may contribute to the antidepressant actions of these agents.

The neurosteroid allopregnanolone is reduced in prefrontal cortex in Alzheimer's disease.

BACKGROUND: Few data are currently available investigating neurosteroids (NS) in Alzheimer's disease (AD). The NS allopregnanolone may be decreased in serum and plasma in patients with AD, but it is unclear if allopregnanolone is also reduced in brain. Because a number of NS exhibit neuroprotective effects and impact cognitive performance in rodent models, these molecules may be relevant to the pathophysiology of neurodegenerative disorders. We therefore investigated prefrontal cortex (PFC) NS levels in AD. METHODS: Neurosteroid levels (allopregnanolone, pregnenolone, dehydroepiandrosterone [DHEA]) were determined in postmortem PFC in 14 male subjects with AD and 15 cognitively intact male control subjects by gas chromatography/mass spectrometry preceded by high-performance liquid chromatography purification. RESULTS: Subjects with AD exhibit significant reductions in allopregnanolone compared with cognitively intact control subjects (median levels = 2.50 ng/g vs. 5.59 ng/g, respectively; p = .02). Allopregnanolone levels are inversely correlated with neuropathological disease stage (Braak), r = -.49, p = .007. Median DHEA levels are elevated in subjects with AD (p = .01). CONCLUSIONS: Subjects with AD demonstrate significant reductions in PFC allopregnanolone levels, a finding that may be relevant to neuropathological disease stage severity. Neurosteroids may have utility as candidate biomarkers in AD.

Neuroactive steroids, negative affect, and nicotine dependence severity in male smokers.

RATIONALE: Nicotine administration alters neuroactive steroids in rodent models, and serum levels of the neuroactive steroid DHEAS (dehydroepiandrosterone sulfate) appear to be higher in smokers. These molecules may be relevant to tobacco addiction and affective symptoms. OBJECTIVES: This study aims to investigate DHEAS, allopregnanolone, pregnenolone, and other steroids in male smokers to determine potential associations with nicotine dependence severity and negative affect. MATERIALS AND METHODS: Allopregnanolone and pregnenolone serum levels were determined by gas chromatography/mass spectrometry, while DHEAS and other steroid levels were determined by radioimmunoassay in 28 male smokers. Correlational analyses were performed to determine potential associations with rating measures, including the Fagerstrom Test for Nicotine Dependence (FTND), the addiction subscale of the Ikard Smoking Motivation Questionnaire (ISMQ), the craving item on the Reasons to Smoke (RTS) Questionnaire, and the negative affect and craving subscales of the Shiffman-Jarvik Withdrawal Questionnaire. RESULTS: DHEAS levels were inversely correlated with the negative affect subscale of the Shiffman-Jarvik Withdrawal Questionnaire (r=-0.60, p=0.002) and the RTS craving item (r=-0.43, p=0.03), and tended to be inversely correlated with the FTND scores (r=-0.38, p=0.067) and the ISMQ addiction subscale (r=-0.38, p=0.059), adjusting for age. Allopregnanolone levels were positively correlated with cotinine levels (r=0.57, p=0.006); pregnenolone levels tended to be positively correlated with cotinine levels (r=0.40, p=0.066). CONCLUSIONS: DHEAS levels were inversely correlated with negative affect and craving measures, and may predict nicotine dependence severity. Allopregnanolone levels were positively correlated with cotinine levels, suggesting that this neuroactive steroid may be upregulated in smokers. Neuroactive steroids may represent novel smoking cessation agents.

Neuroactive steroids are altered in schizophrenia and bipolar disorder: relevance to pathophysiology and therapeutics.

Evidence suggests that neuroactive steroids may be candidate modulators of schizophrenia pathophysiology and therapeutics. We therefore investigated neuroactive steroid levels in post-mortem brain tissue from subjects with schizophrenia, bipolar disorder, nonpsychotic depression, and control subjects to determine if neuroactive steroids are altered in these disorders. Posterior cingulate and parietal cortex tissue from the Stanley Foundation Neuropathology Consortium collection was analyzed for neuroactive steroids by negative ion chemical ionization gas chromatography/mass spectrometry preceded by high-performance liquid chromatography. Subjects with schizophrenia, bipolar disorder, nonpsychotic depression, and control subjects were group matched for age, sex, ethnicity, brain pH, and post-mortem interval (n = 14-15 per group, 59-60 subjects total). Statistical analyses were performed by ANOVA with post-hoc Dunnett tests on log transformed neuroactive steroid levels. Pregnenolone and allopregnanolone were present in human post-mortem brain tissue at considerably higher concentrations than typically observed in serum or plasma. Pregnenolone and dehydroepiandrosterone levels were higher in subjects with schizophrenia and bipolar disorder compared to control subjects in both posterior cingulate and parietal cortex. Allopregnanolone levels tended to be decreased in parietal cortex in subjects with schizophrenia compared to control subjects. Neuroactive steroids are present in human post-mortem brain tissue at physiologically relevant concentrations and altered in subjects with schizophrenia and bipolar disorder. A number of neuroactive steroids act at inhibitory GABA(A) and excitatory NMDA receptors and demonstrate neuroprotective and neurotrophic effects. Neuroactive steroids may therefore be candidate modulators of the pathophysiology of schizophrenia and bipolar disorder, and relevant to the treatment of these disorders.

Apolipoprotein E genotype and subcortical vascular lesions in older depressed patients and control subjects.

BACKGROUND: Several studies have linked geriatric depression with cerebrovascular disease. The apolipoprotein E gene (APOE) epsilon 4 allele has been associated with a variety of late-life neuropsychiatric disorders, including Alzheimer's disease, vascular dementia, and depression. METHODS: The sample consisted of 145 elderly depressive individuals and 100 nondepressed elderly control subjects. After a standardized clinical assessment, all subjects underwent a magnetic resonance imaging brain scan. Volumes of subcortical white and gray matter lesions were determined using a semi-automated method. Apolipoprotein E genotype was determined on blood sample using a standard protocol. A series of linear regression models were developed to assess the relationships between APOE genotype and white and gray matter lesion volumes. RESULTS: Older age, lower Mini-Mental State Examination score, and having any APOE epsilon 4 allele were each correlated with gray-matter lesion volume in depressed patients. Apolipoprotein E genotype was not associated with any lesion volume among control subjects. In a subsequent linear regression model, gray matter lesion volume was associated with older age, having at least one APOE epsilon 4 allele, and white matter lesion volume among depressed patients. CONCLUSIONS: These results are consistent with previous reports linking cerebrovascular disease and APOE genotype. Further studies are needed to replicate this finding in elderly depressive individuals and to explain the relationship between the APOE locus and development of central nervous system vascular pathology.