Abstract
The introduction of mild cognitive impairment (MCI) as a diagnostic category adds to the challenges of diagnosing Alzheimer's Disease (AD). No single marker has been proven to accurately categorize patients into their respective diagnostic groups. Thus, previous studies have attempted to develop fused predictors of AD and MCI. These studies have two main limitations. Most do not simultaneously consider all diagnostic categories and provide suboptimal fused representations using the same set of modalities for prediction of all classes. In this work, we present a combined framework, cascaded multiview canonical correlation (CaMCCo), for fusion and cascaded classification that incorporates all diagnostic categories and optimizes classification by selectively combining a subset of modalities at each level of the cascade. CaMCCo is evaluated on a data cohort comprising 149 patients for whom neurophysiological, neuroimaging, proteomic and genomic data were available. Results suggest that fusion of select modalities for each classification task outperforms (mean AUC = 0.92) fusion of all modalities (mean AUC = 0.54) and individual modalities (mean AUC = 0.90, 0.53, 0.71, 0.73, 0.62, 0.68). In addition, CaMCCo outperforms all other multi-class classification methods for MCI prediction (PPV: 0.80 vs. 0.67, 0.63).
Original language | English |
---|---|
Article number | 8137 |
Journal | Scientific Reports |
Volume | 7 |
Issue number | 1 |
DOIs | |
State | Published - 1 Dec 2017 |
Fingerprint
Dive into the research topics of 'Cascaded Multi-view Canonical Correlation (CaMCCo) for Early Diagnosis of Alzheimer's Disease via Fusion of Clinical, Imaging and Omic Features'. Together they form a unique fingerprint.Cite this
- APA
- Author
- BIBTEX
- Harvard
- Standard
- RIS
- Vancouver
}
In: Scientific Reports, Vol. 7, No. 1, 8137, 01.12.2017.
Research output: Contribution to journal › Article › peer-review
TY - JOUR
T1 - Cascaded Multi-view Canonical Correlation (CaMCCo) for Early Diagnosis of Alzheimer's Disease via Fusion of Clinical, Imaging and Omic Features
AU - Singanamalli, Asha
AU - Wang, Haibo
AU - Madabhushi, Anant
AU - Weiner, Michael
AU - Aisen, Paul
AU - Petersen, Ronald
AU - Jack, Clifford
AU - Jagust, William
AU - Trojanowki, John
AU - Toga, Arthur
AU - Beckett, Laurel
AU - Green, Robert
AU - Saykin, Andrew
AU - Morris, John
AU - Shaw, Leslie
AU - Kaye, Jeffrey
AU - Quinn, Joseph
AU - Silbert, Lisa
AU - Lind, Betty
AU - Carter, Raina
AU - Dolen, Sara
AU - Schneider, Lon
AU - Pawluczyk, Sonia
AU - Beccera, Mauricio
AU - Teodoro, Liberty
AU - Spann, Bryan
AU - Brewer, James
AU - Vanderswag, Helen
AU - Fleisher, Adam
AU - Heidebrink, Judith
AU - Lord, Joanne
AU - Mason, Sara
AU - Albers, Colleen
AU - Knopman, David
AU - Johnson, Kris
AU - Doody, Rachelle
AU - Villanueva-Meyer, Javier
AU - Chowdhury, Munir
AU - Rountree, Susan
AU - Dang, Mimi
AU - Stern, Yaakov
AU - Honig, Lawrence
AU - Bell, Karen
AU - Ances, Beau
AU - Carroll, Maria
AU - Creech, Mary
AU - Franklin, Erin
AU - Mintun, Mark
AU - Schneider, Stacy
AU - Oliver, Angela
AU - Marson, Daniel
AU - Griffith, Randall
AU - Clark, David
AU - Geldmacher, David
AU - Brockington, John
AU - Roberson, Erik
AU - Natelson Love, Marissa
AU - Grossman, Hillel
AU - Mitsis, Effie
AU - Shah, Raj
AU - Detoledo-Morrell, Leyla
AU - Duara, Ranjan
AU - Varon, Daniel
AU - Greig, Maria
AU - Roberts, Peggy
AU - Albert, Marilyn
AU - Onyike, Chiadi
AU - D'Agostino, Daniel
AU - Kielb, Stephanie
AU - Galvin, James
AU - Cerbone, Brittany
AU - Michel, Christina
AU - Pogorelec, Dana
AU - Rusinek, Henry
AU - De Leon, Mony
AU - Glodzik, Lidia
AU - De Santi, Susan
AU - Doraiswamy, P.
AU - Petrella, Jeffrey
AU - Borges-Neto, Salvador
AU - Wong, Terence
AU - Coleman, Edward
AU - Smith, Charles
AU - Jicha, Greg
AU - Hardy, Peter
AU - Sinha, Partha
AU - Oates, Elizabeth
AU - Conrad, Gary
AU - Porsteinsson, Anton
AU - Goldstein, Bonnie
AU - Martin, Kim
AU - Makino, Kelly
AU - Ismail, M.
AU - Brand, Connie
AU - Mulnard, Ruth
AU - Thai, Gaby
AU - Mc-Adams-Ortiz, Catherine
AU - Womack, Kyle
AU - Mathews, Dana
AU - Quiceno, Mary
AU - Levey, Allan
AU - Lah, James
AU - Cellar, Janet
AU - Burns, Jeffrey
AU - Swerdlow, Russell
AU - Brooks, William
AU - Apostolova, Liana
AU - Tingus, Kathleen
AU - Woo, Ellen
AU - Silverman, Daniel
AU - Lu, Po
AU - Bartzokis, George
AU - Graff-Radford, Neill
AU - Parfitt, Francine
AU - Kendall, Tracy
AU - Johnson, Heather
AU - Farlow, Martin
AU - Marie Hake, Ann
AU - Matthews, Brandy
AU - Brosch, Jared
AU - Herring, Scott
AU - Hunt, Cynthia
AU - Dyck, Christopher
AU - Carson, Richard
AU - MacAvoy, Martha
AU - Varma, Pradeep
AU - Chertkow, Howard
AU - Bergman, Howard
AU - Hosein, Chris
AU - Black, Sandra
AU - Stefanovic, Bojana
AU - Caldwell, Curtis
AU - Robin Hsiung, Ging Yuek
AU - Feldman, Howard
AU - Mudge, Benita
AU - Assaly, Michele
AU - Finger, Elizabeth
AU - Pasternack, Stephen
AU - Rachisky, Irina
AU - Trost, Dick
AU - Kertesz, Andrew
AU - Bernick, Charles
AU - Munic, Donna
AU - Mesulam, Marek Marsel
AU - Lipowski, Kristine
AU - Weintraub, Sandra
AU - Bonakdarpour, Borna
AU - Kerwin, Diana
AU - Wu, Chuang Kuo
AU - Johnson, Nancy
AU - Sadowsky, Carl
AU - Villena, Teresa
AU - Scott Turner, Raymond
AU - Johnson, Kathleen
AU - Reynolds, Brigid
AU - Sperling, Reisa
AU - Johnson, Keith
AU - Marshall, Gad
AU - Yesavage, Jerome
AU - Taylor, Joy
AU - Lane, Barton
AU - Rosen, Allyson
AU - Tinklenberg, Jared
AU - Sabbagh, Marwan
AU - Belden, Christine
AU - Jacobson, Sandra
AU - Sirrel, Sherye
AU - Kowall, Neil
AU - Killiany, Ronald
AU - Budson, Andrew
AU - Norbash, Alexander
AU - Lynn Johnson, Patricia
AU - Obisesan, Thomas
AU - Wolday, Saba
AU - Allard, Joanne
AU - Lerner, Alan
AU - Ogrocki, Paula
AU - Tatsuoka, Curtis
AU - Fatica, Parianne
AU - Fletcher, Evan
AU - Maillard, Pauline
AU - Olichney, John
AU - Decarli, Charles
AU - Carmichael, Owen
AU - Kittur, Smita
AU - Borrie, Michael
AU - Lee, T. Y.
AU - Robbartha,
AU - Johnson, Sterling
AU - Asthana, Sanjay
AU - Carlsson, Cynthia
AU - Potkin, Steven
AU - Preda, Adrian
AU - Nguyen, Dana
AU - Tariot, Pierre
AU - Burke, Anna
AU - Trncic, Nadira
AU - Reeder, Stephanie
AU - Bates, Vernice
AU - Capote, Horacio
AU - Rainka, Michelle
AU - Scharre, Douglas
AU - Kataki, Maria
AU - Adeli, Anahita
AU - Zimmerman, Earl
AU - Celmins, Dzintra
AU - Brown, Alice
AU - Pearlson, Godfrey
AU - Blank, Karen
AU - Anderson, Karen
AU - Flashman, Laura
AU - Seltzer, Marc
AU - Hynes, Mary
AU - Santulli, Robert
AU - Sink, Kaycee
AU - Gordineer, Leslie
AU - Williamson, Jeff
AU - Garg, Pradeep
AU - Watkins, Franklin
AU - Ott, Brian
AU - Querfurth, Henry
AU - Tremont, Geoffrey
AU - Salloway, Stephen
AU - Malloy, Paul
AU - Correia, Stephen
AU - Rosen, Howard
AU - Miller, Bruce
AU - Perry, David
AU - Mintzer, Jacobo
AU - Spicer, Kenneth
AU - Bachman, David
AU - Pomara, Nunzio
AU - Hernando, Raymundo
AU - Sarrael, Antero
AU - Relkin, Norman
AU - Chaing, Gloria
AU - Lin, Michael
AU - Ravdin, Lisa
AU - Smith, Amanda
AU - Ashok Raj, Balebail
AU - Fargher, Kristin
N1 - Publisher Copyright: © 2017 The Author(s).
PY - 2017/12/1
Y1 - 2017/12/1
N2 - The introduction of mild cognitive impairment (MCI) as a diagnostic category adds to the challenges of diagnosing Alzheimer's Disease (AD). No single marker has been proven to accurately categorize patients into their respective diagnostic groups. Thus, previous studies have attempted to develop fused predictors of AD and MCI. These studies have two main limitations. Most do not simultaneously consider all diagnostic categories and provide suboptimal fused representations using the same set of modalities for prediction of all classes. In this work, we present a combined framework, cascaded multiview canonical correlation (CaMCCo), for fusion and cascaded classification that incorporates all diagnostic categories and optimizes classification by selectively combining a subset of modalities at each level of the cascade. CaMCCo is evaluated on a data cohort comprising 149 patients for whom neurophysiological, neuroimaging, proteomic and genomic data were available. Results suggest that fusion of select modalities for each classification task outperforms (mean AUC = 0.92) fusion of all modalities (mean AUC = 0.54) and individual modalities (mean AUC = 0.90, 0.53, 0.71, 0.73, 0.62, 0.68). In addition, CaMCCo outperforms all other multi-class classification methods for MCI prediction (PPV: 0.80 vs. 0.67, 0.63).
AB - The introduction of mild cognitive impairment (MCI) as a diagnostic category adds to the challenges of diagnosing Alzheimer's Disease (AD). No single marker has been proven to accurately categorize patients into their respective diagnostic groups. Thus, previous studies have attempted to develop fused predictors of AD and MCI. These studies have two main limitations. Most do not simultaneously consider all diagnostic categories and provide suboptimal fused representations using the same set of modalities for prediction of all classes. In this work, we present a combined framework, cascaded multiview canonical correlation (CaMCCo), for fusion and cascaded classification that incorporates all diagnostic categories and optimizes classification by selectively combining a subset of modalities at each level of the cascade. CaMCCo is evaluated on a data cohort comprising 149 patients for whom neurophysiological, neuroimaging, proteomic and genomic data were available. Results suggest that fusion of select modalities for each classification task outperforms (mean AUC = 0.92) fusion of all modalities (mean AUC = 0.54) and individual modalities (mean AUC = 0.90, 0.53, 0.71, 0.73, 0.62, 0.68). In addition, CaMCCo outperforms all other multi-class classification methods for MCI prediction (PPV: 0.80 vs. 0.67, 0.63).
UR - http://www.scopus.com/inward/record.url?scp=85027495943&partnerID=8YFLogxK
U2 - 10.1038/s41598-017-03925-0
DO - 10.1038/s41598-017-03925-0
M3 - Article
C2 - 28811553
AN - SCOPUS:85027495943
SN - 2045-2322
VL - 7
JO - Scientific Reports
JF - Scientific Reports
IS - 1
M1 - 8137
ER -