Wolz 2011.
| Study characteristics | |||
| Patient sampling |
Primary objectives: to assess the improvement in classification accuracy that can be achieved by combining features from different structural MRI analysis techniques (hippocampal volume, tensor‐based morphometry, cortical thickness and a novel technique based on manifold learning). Study population: participants with MCI included in ADNI for which a 1.5 T T1‐weighted MRI scan at baseline was available Selection criteria: MCI included participants had MMSE scores between 24–30 (inclusive), a memory complaint, objective memory loss measured by education‐adjusted scores on WMS Logical Memory II, a CDR of 0.5, absence of significant levels of impairment in other cognitive domains, essentially preserved ADL, and an absence of dementia. Exclusion criteria not reported in the published article. Exclusion criteria as reported in the ADNI protocol ADNI 2010 and Gaser 2013 Study design: prospective longitudinal study (ADNI study) |
||
| Patient characteristics and setting |
Clinical presentations: amnestic MCI according to ADNI 2010 and Petersen 2010 Age mean (SD): MCI who progressed to AD: 75 ± 7; stable MCI: 75 ± 8 Gender (% men): MCI who progressed to AD: 62%; stable MCI: 66% Education years mean (SD): MCI who progressed to AD: 15.7 ± 2.9; stable MCI: 15.6 ± 3.1 ApoE4 carriers (%): MCI who progressed to AD: 66%; stable MCI: 39% Neuropsychological tests: MMSE mean (SD): MCI who progressed to AD: 26.6 ± 1.7; stable MCI 27.3 ± 1.8 Clinical stroke excluded: not specified Co‐morbidities: not reported Number enrolled: 405 Number available for analysis: 405 Setting: participants of the ADNI study Country: USA and Canada Period: follow‐up period was stopped in July 2011 Language: English |
||
| Index tests |
Index test: MRI automated method for estimation of hippocampal volume; also other methods described Manufacturer: GE Healthcare, Philips Medical System, Siemens Medical Solution (Jack 2008b) Tesla strength: 1.5 Assessment methods: hippocampal volumes were measured using an approach based on fast and robust multi‐atlas segmentation (Lötjönen 2011). Details about cortical thickness, tensor‐based morphometry and minifold‐based learning methods were also reported. 2 different methods were used to perform classification based on individual features and their combination: linear discriminant analysis and support vector machines. Total volume of left and right hippocampus were combined in a single feature. Description of positive cases definition by index test as reported: classifiers were built on a training set composed of healthy and AD and used to classify the MRI images of MCI participants as more similar to healthy (negative cases) and more similar to AD (positive cases) . Examiners: imaging interpretation reserved to an automatic classifier Interobserver variability: not provided |
||
| Target condition and reference standard(s) |
Target condition: AD Prevalence of AD in the sample: 167/405 (41% of enrolled participants) Stable MCI or converted to other dementia: 238 (59%) stable MCI Reference standards: NINCDS‐ADRDA criteria (McKhann 1984). Mean clinical follow‐up: 1.5 years |
||
| Flow and timing |
Withdrawals explained and losses to follow‐up: none reported Uninterpretable MRI results have not been reported |
||
| Comparative | |||
| Key conclusions by the authors | A comprehensive analysis of MRI images combining multiple features improves classification accuracy and predictive power in detecting early AD. The most stable and reliable classification was achieved when combining all available features. | ||
| Conflict of interests | The study authors have declared no competing interests | ||
| Notes |
Source of funding: U01 AG024904 2 x 2 table: data from the published article; we only used data regarding volumetric results obtained with linear discriminant analysis in the review |
||
| Methodological quality | |||
| Item | Authors' judgement | Risk of bias | Applicability concerns |
| DOMAIN 1: Patient Selection | |||
| Was a consecutive or random sample of patients enrolled? | No | ||
| Was a case‐control design avoided? | Yes | ||
| Did the study avoid inappropriate exclusions? | No | ||
| High | Low | ||
| DOMAIN 2: Index Test All tests | |||
| Were the index test results interpreted without knowledge of the results of the reference standard? | Unclear | ||
| Did the study provide a clear pre‐specified definition of what was considered to be a "positive" result of the index test? | Yes | ||
| Was the index test performed by a single operator or interpreted by consensus in a joint session? | Unclear | ||
| Unclear | Low | ||
| DOMAIN 3: Reference Standard | |||
| Is the reference standards likely to correctly classify the target condition? | Yes | ||
| Were the reference standard results interpreted without knowledge of the results of the index tests? | Yes | ||
| Low | Low | ||
| DOMAIN 4: Flow and Timing | |||
| Was there an appropriate interval between index test and reference standard? | Yes | ||
| Did all patients receive the same reference standard? | Yes | ||
| Were all patients included in the analysis? | Yes | ||
| Low | |||